Docs/MCP Tool Catalog
Live Tool Schemas47 Connectors · 543 Tools

MCP Tool Catalog

Nunc includes 47 managed MCP connectors exposing 502 tools, plus 41 built-in platform tools. Every schema below matches the live tools/list definitions served to authenticated clients.

How tools are exposed across Nunc endpoints

Nunc Tools MCP (https://api.nunc.cloud/mcp): Serves enabled connectors and built-in platform tools to Claude, ChatGPT, Gemini, Cursor, and VS Code (append ?apps=slack,salesforce to focus exposed schemas). Every call enforces 3-tier PBAC, Guardian PII masking, and human approval gates.

Nunc Memory MCP (https://api.nunc.cloud/mcp/memory) & Nunc Sandbox MCP (/mcp/sandbox): Dedicated servers for four-tier persistent memory (Public, Company, Workspace, Agent — pinned optionally via ?workspace=<id>) and isolated Python/JS/TS sandbox execution.

Nunc Delegate MCP (https://api.nunc.cloud/mcp/delegate) & Nunc Analytics MCP (/mcp/analytics): Per-user governed agent delegation with loop guards and remote cancellation, plus zero-copy warehouse analytics (Analytics MCP) for Data Studio.

Nunc Service API (https://api.nunc.cloud/service): Headless JSON-RPC endpoint restricted to service application keys for deterministic backend data syncs with full policy, rate-limit, and audit enforcement.

communication (8)

Gmail gmail · 12 tools · Per-user OAuth / credentials

Gmail via delegated Google OAuth (users/me). Connect from Desktop — not a shared workspace mailbox.

Transport: cloud + local (stdio) · Required credentials: GMAIL_CLIENT_ID, GMAIL_CLIENT_SECRET, GMAIL_ACCESS_TOKEN

ToolParametersReturnsUser Scoping
gmail_list_messages
List emails matching a query using Gmail search syntax
  • query?: string — Gmail search query (e.g. "is:unread", "from:alice@example.com")
  • maxResults?: number — Maximum number of messages to return (default 10, max 500)
  • labelIds?: array<string> — Only return messages with these label IDs
JSON passthrough from provider API—
gmail_get_message
Get the full content of an email message
  • messageId: string — The message ID
  • format?: string — Response format
JSON passthrough from provider API—
gmail_send_message
Send a new email message via the user's Gmail account with optional HTML body and recipients.
  • to: string — Recipient email address(es), comma-separated
  • subject: string — Email subject line
  • body: string — Plain text or HTML email body
  • cc?: string — CC recipients, comma-separated
  • bcc?: string — BCC recipients, comma-separated
  • replyToMessageId?: string — Message ID to set as In-Reply-To (for threading)
  • isHtml?: boolean — Set to true if body is HTML content
JSON passthrough from provider API—
gmail_reply
Reply to an existing email message (maintains thread)
  • messageId: string — The message ID to reply to
  • body: string — Plain text or HTML reply body
  • isHtml?: boolean — Set to true if body is HTML content
JSON passthrough from provider API—
gmail_forward
Forward an email message to another recipient
  • messageId: string — The message ID to forward
  • to: string — Recipient email address(es), comma-separated
JSON passthrough from provider API—
gmail_trash_message
Move an email message to the trash
  • messageId: string — The message ID to trash
JSON passthrough from provider API—
gmail_label_message
Add or remove labels on an email message
  • messageId: string — The message ID
  • addLabelIds?: array<string> — Label IDs to add
  • removeLabelIds?: array<string> — Label IDs to remove
JSON passthrough from provider API—
gmail_list_labels
List all labels in the Gmail account
noneJSON passthrough from provider API—
gmail_search
Full-text search across emails using Gmail search syntax
  • query: string — Gmail search query
  • maxResults?: number — Maximum results to return (default 10)
JSON passthrough from provider API—
gmail_get_thread
Get all messages in an email thread
  • threadId: string — The thread ID
JSON passthrough from provider API—
gmail_list_drafts
List all draft emails currently stored in the user's Gmail mailbox.
noneJSON passthrough from provider API—
gmail_create_draft
Create a new draft email message in Gmail without sending it immediately.
  • to: string — Recipient email address(es), comma-separated
  • subject: string — Email subject line
  • body: string — Plain text or HTML email body
  • isHtml?: boolean — Set to true if body is HTML content
JSON passthrough from provider API—
Gmail Hosted MCP gmail-mcp · runtime-discovered · Per-user OAuth / credentials

Federates Google's official Gmail MCP (gmailmcp.googleapis.com) through Nunc /mcp. Each Nunc user Connects Google as themselves; tools inherit that mailbox. Draft, search, and label only — first-party gmail_send_message / reply / forward remain the send path.

Transport: cloud

Tools are discovered dynamically at runtime via tools/list once connected.

Meltwater Hosted MCP meltwater-mcp · runtime-discovered · Per-user OAuth / credentials

Federates Meltwater's hosted MCP (api.meltwater.com/v2/mcp) through Nunc /mcp. Access Meltwater news and social mentions, analytics, and saved searches. Supports Meltwater API tokens (apikey header) and per-user OAuth.

Transport: cloud · Required credentials: MELTWATER_API_KEY

Tools are discovered dynamically at runtime via tools/list once connected.

Meltwater (Service) meltwater-service · 5 tools · Organization credentials

Server-to-server Meltwater integration for autonomous applications, briefings, and media monitoring pipelines. Uses an organization API token with mw_svc_* tools so PBAC can grant background media intelligence to service keys while keeping interactive user credentials separate.

Transport: cloud · Required credentials: MELTWATER_SVC_API_KEY, MELTWATER_SVC_BASE_URL

ToolParametersReturnsUser Scoping
mw_svc_search_mentions
Search news and social media mentions across Meltwater's media intelligence database. Filter by query keywords, date range, sentiment, source type, or saved search ID.
  • query?: string — Search query keywords or boolean expression (e.g. 'Apple AND iPhone NOT refurb').
  • search_id?: string — Optional saved search ID to query directly.
  • source_types?: array<string> — Types of sources to include (e.g. ['news', 'blogs', 'twitter', 'reddit', 'broadcast']).
  • start_date?: string — Start date in ISO-8601 format (e.g. '2026-09-01T00:00:00Z').
  • end_date?: string — End date in ISO-8601 format (e.g. '2026-09-15T23:59:59Z').
  • sentiment?: string — Filter mentions by analyzed sentiment.
  • limit?: number — Maximum number of mentions to return (default 25, max 100).
JSON passthrough from provider API—
mw_svc_get_mention_detail
Retrieve full article text, metadata, reach, author, and sentiment details for a specific Meltwater mention by ID.
  • mention_id: string — Unique Meltwater mention / document ID.
JSON passthrough from provider API—
mw_svc_get_saved_searches
List all configured Meltwater saved searches and topic monitors in the organization.
  • limit?: number — Maximum number of saved searches to list (default 50).
JSON passthrough from provider API—
mw_svc_get_analytics
Retrieve aggregated media intelligence analytics, share-of-voice, volume trends, sentiment breakdown, and top publications for a query or search ID.
  • search_id?: string — Saved search ID to calculate analytics for.
  • query?: string — Ad-hoc query expression if no search_id is specified.
  • metric?: string — Analytical metric to compute (default: 'volume').
  • start_date?: string — Start date in ISO-8601 format.
  • end_date?: string — End date in ISO-8601 format.
  • interval?: string — Time bucket interval for volume/trend metrics (default: 'day').
JSON passthrough from provider API—
mw_svc_list_tags
List tags and categories used for content curation and document categorization across the organization.
  • limit?: number — Maximum number of tags to return (default 50).
JSON passthrough from provider API—
Outlook outlook · 23 tools · Per-user OAuth / credentials

Microsoft Outlook Mail, Calendar, and Contacts via delegated Microsoft Graph (/me). People-lane user OAuth — not application permissions.

Transport: cloud + local (stdio) · Required credentials: MICROSOFT_TENANT_ID, MICROSOFT_CLIENT_ID, MICROSOFT_CLIENT_SECRET, MICROSOFT_ACCESS_TOKEN

ToolParametersReturnsUser Scoping
outlook_list_messages
Microsoft Outlook Mail — List email messages with optional folder scoping (including nested subfolders), filtering, pagination, and search. Graph forbids combining $search and $filter; page via @odata.nextLink.
  • folderId?: string — Optional mail folder ID, well-known folder name ('inbox', 'sentitems', 'drafts', 'deleteditems'), subfolder display name (e.g. 'Projects'), or folder path (e.g. 'Inbox/Projects/Q4'). If unsure of the exact folder ID or name, call outlook_list_folders first. Omit to list across the mailbox.
  • filter?: string — OData $filter expression (e.g. "isRead eq false" or "receivedDateTime ge 2026-11-01T00:00:00Z"). Note: Microsoft Graph does NOT support contains() on email addresses or complex extended properties in $filter; use search or outlook_search_messages for keyword/sender searches. Cannot be combined with search.
  • top?: number — Number of messages to return (max 50)
  • skip?: number — Number of messages to skip for pagination
  • search?: string — Search query string. Cannot be combined with filter — use outlook_search_messages or omit filter.
  • select?: string — Comma-separated fields to select
JSON passthrough from provider API—
outlook_get_message
Microsoft Outlook Mail — Get a specific email message by ID with full details.
  • id: string — Unique Microsoft Graph message identifier string (from an outlook_list_messages or outlook_search_messages result). Do not pass folder names or subject lines.
JSON passthrough from provider API—
outlook_send_message
Microsoft Outlook Mail — Send a new email message via Microsoft Graph Outlook with optional HTML body.
  • to: array<string> — Recipient email addresses
  • subject: string — Email subject
  • body: string — Email body (HTML supported)
  • cc?: array<string> — CC email addresses
  • bcc?: array<string> — BCC email addresses
  • importance?: string — Message importance
  • attachments?: array<object> — Optional list of document files or text attachments to include with the email message.
JSON passthrough from provider API—
outlook_reply
Microsoft Outlook Mail — Reply to an email message.
  • id: string — Message ID to reply to
  • comment: string — Reply body text (HTML supported)
JSON passthrough from provider API—
outlook_forward
Microsoft Outlook Mail — Forward an email message to recipients.
  • id: string — Message ID to forward
  • to: array<string> — Recipient email addresses
  • comment?: string — Comment to include with forwarded message
JSON passthrough from provider API—
outlook_move_message
Microsoft Outlook Mail — Move an email message to a different mail folder.
  • id: string — Unique Microsoft Graph message identifier string to move
  • destinationId: string — Destination folder ID, well-known folder name (e.g. 'archive', 'deleteditems'), subfolder display name, or folder path (e.g. 'Inbox/Projects/Q4')
JSON passthrough from provider API—
outlook_delete_message
Microsoft Outlook Mail — Permanently delete an email message from Outlook by message ID.
  • id: string — Message ID to delete
JSON passthrough from provider API—
outlook_list_folders
Microsoft Outlook Mail — List mail folders and nested subfolders (Inbox, Sent, Drafts, Archive, and custom subfolder trees). Optionally pass parentFolderId (ID, well-known name, subfolder name, or path like 'Inbox/Projects') to list subfolders of a specific folder.
  • parentFolderId?: string — Optional parent folder ID, well-known folder name (e.g. 'inbox'), folder display name, or folder path (e.g. 'Inbox/Projects' or 'Inbox -> Projects') to list subfolders of. Omit to list all folders and nested subfolders.
  • includeChildFolders?: boolean — When true (default), recursively expands nested childFolders and includes each folder's full path (e.g. 'Inbox/Projects/Q4').
JSON passthrough from provider API—
outlook_search_messages
Microsoft Outlook Mail — Search email messages using a keyword query across the mailbox or within a specific folder/subfolder.
  • query: string — Search query string
  • folderId?: string — Optional mail folder ID, well-known folder name (e.g. 'inbox'), subfolder display name (e.g. 'Projects'), or folder path (e.g. 'Inbox/Projects/Q4') to restrict the search to a specific folder. If unsure of exact folder name or ID, call outlook_list_folders first. Omit to search the entire mailbox.
  • top?: number — Number of results to return
JSON passthrough from provider API—
outlook_create_draft
Microsoft Outlook Mail — Create a draft email message.
  • to?: array<string> — Recipient email addresses
  • subject: string — Email subject
  • body: string — Email body (HTML supported)
  • cc?: array<string> — CC email addresses
  • bcc?: array<string> — BCC email addresses
  • importance?: string — Message importance level (low, normal, high)
  • attachments?: array<object> — Optional list of document files or text attachments to include with the email message.
JSON passthrough from provider API—
outlook_list_attachments
Microsoft Outlook Mail — List attachments for a specific email message.
  • id: string — Unique Microsoft Graph message identifier whose attachments should be listed
JSON passthrough from provider API—
outlook_flag_message
Microsoft Outlook Mail — Set or clear a follow-up flag on an email message.
  • id: string — Unique Microsoft Graph message identifier string (from an outlook_list_messages or outlook_search_messages result). Do not pass folder names or subject lines.
  • flagStatus: string — Follow-up status flag: 'flagged' marks message for follow-up, 'complete' marks completed, 'notFlagged' clears existing flag
JSON passthrough from provider API—
outlook_list_events
Microsoft Outlook Calendar — List calendar events in a required start/end window via GET /me/calendarView (not /me/events). Default calendar only.
  • start: string — Start of range (ISO 8601). Required Graph calendarView startDateTime.
  • end: string — End of range (ISO 8601). Required Graph calendarView endDateTime.
  • top?: number — Number of events to return (max 50)
  • filter?: string — OData $filter expression
  • select?: string — Comma-separated fields to select
  • timeZone?: string — Optional Windows or IANA time zone for returned start/end times (e.g. 'America/New_York')
JSON passthrough from provider API—
outlook_get_event
Microsoft Outlook Calendar — Get a calendar event by ID (GET /me/events/{id}). Default calendar only.
  • id: string — Unique Microsoft Graph calendar event identifier string
  • timeZone?: string — Optional Windows or IANA time zone for returned start/end times (e.g. 'America/New_York')
JSON passthrough from provider API—
outlook_create_event
Microsoft Outlook Calendar — Schedule / create a calendar event (POST /me/events). Default calendar only.
  • subject: string — Event title
  • start: string — Start datetime (ISO 8601)
  • end: string — End datetime (ISO 8601)
  • body?: string — Event body (HTML supported)
  • location?: string — Location display name
  • attendees?: array<string> — Attendee email addresses
  • isAllDay?: boolean — All-day event
  • timeZone?: string — Windows or IANA time zone for start/end (default UTC)
JSON passthrough from provider APIparam: `attendees` (caller-asserted)
outlook_update_event
Microsoft Outlook Calendar — Update an existing calendar event (PATCH /me/events/{id}).
  • id: string — Unique Microsoft Graph calendar event identifier string
  • subject?: string — Updated event title
  • start?: string — Updated start datetime (ISO 8601)
  • end?: string — Updated end datetime (ISO 8601)
  • body?: string — Updated event body (HTML supported)
  • location?: string — Updated location display name
  • attendees?: array<string> — Updated attendee email addresses
  • isAllDay?: boolean — All-day event
  • timeZone?: string — Windows or IANA time zone for start/end
JSON passthrough from provider APIparam: `attendees` (caller-asserted)
outlook_delete_event
Microsoft Outlook Calendar — Delete / cancel a calendar event (DELETE /me/events/{id}).
  • id: string — Event ID to delete
JSON passthrough from provider API—
outlook_list_contacts
Microsoft Outlook Contacts — List contacts from the user's Outlook address book (GET /me/contacts) with optional filtering, search, and pagination.
  • filter?: string — OData $filter expression (e.g. "companyName eq 'Acme'"). Cannot be combined with search.
  • top?: number — Number of contacts to return (max 50)
  • skip?: number — Number of contacts to skip for pagination
  • search?: string — Search query string. Cannot be combined with filter — use outlook_search_contacts or omit filter.
  • select?: string — Comma-separated fields to select
JSON passthrough from provider API—
outlook_get_contact
Microsoft Outlook Contacts — Get a specific contact by ID from the user's Outlook address book (GET /me/contacts/{id}).
  • id: string — Unique Microsoft Graph contact identifier string
JSON passthrough from provider API—
outlook_search_contacts
Microsoft Outlook Contacts — Search contacts by name, email address, phone, or company (GET /me/contacts?$search).
  • query: string — Search query string (name, email, company, etc.)
  • top?: number — Number of contacts to return (max 50)
JSON passthrough from provider API—
outlook_create_contact
Microsoft Outlook Contacts — Create a new contact in the user's Outlook address book (POST /me/contacts).
  • givenName: string — Contact's first name (e.g. 'Alice')
  • surname?: string — Contact's last or family name (e.g. 'Smith')
  • displayName?: string — Full formatted name shown in the address book
  • emailAddresses?: array<string> — List of email addresses for the contact
  • businessPhones?: array<string> — List of office or work telephone numbers
  • mobilePhone?: string — Cellular or mobile telephone number for the contact
  • companyName?: string — Organization or employer name
  • jobTitle?: string — Professional role or position at the organization
  • department?: string — Organizational unit or team within the company
  • personalNotes?: string — Free-text notes or context stored on the contact card
JSON passthrough from provider API—
outlook_update_contact
Microsoft Outlook Contacts — Update an existing contact in the user's Outlook address book (PATCH /me/contacts/{id}).
  • id: string — Unique Microsoft Graph contact identifier string
  • givenName?: string — Updated first name for the contact
  • surname?: string — Updated last or family name for the contact
  • displayName?: string — Updated full formatted name shown in the address book
  • emailAddresses?: array<string> — Replacement list of email addresses for the contact
  • businessPhones?: array<string> — Replacement list of office or work telephone numbers
  • mobilePhone?: string — Updated cellular or mobile telephone number
  • companyName?: string — Updated organization or employer name
  • jobTitle?: string — Updated professional role or position at the organization
  • department?: string — Updated organizational unit or team within the company
  • personalNotes?: string — Updated free-text notes stored on the contact card
JSON passthrough from provider API—
outlook_delete_contact
Microsoft Outlook Contacts — Delete a contact from the user's Outlook address book (DELETE /me/contacts/{id}).
  • id: string — Unique Microsoft Graph contact identifier string to delete
JSON passthrough from provider API—
Slack slack · 15 tools · Per-user OAuth / credentials

Slack workspace integration for messaging, channel management, user lookups, and thread search. Delegated user OAuth or bot token authentication.

Transport: cloud + local (stdio) · Required credentials: SLACK_BOT_TOKEN, SLACK_USER_TOKEN, SLACK_TEAM_ID, CLIENT_ID, CLIENT_SECRET

ToolParametersReturnsUser Scoping
slack_list_channels
List public and private Slack channels that the bot has access to. Use to discover conversation IDs for posting or reading history. Returns channel IDs, names, topics, and member counts.
  • types?: string — Comma-separated channel types to include: 'public_channel', 'private_channel', 'mpim' (group DM), or 'im' (direct message)
  • limit?: number — Maximum number of channels to return in this page (min 1, max 1000, default 100)
JSON passthrough from provider API—
slack_get_channel_info
Retrieve detailed metadata for a Slack channel including topic, purpose description, member count, and creator ID. Use before posting to verify channel context.
  • channelId: string — Alphanumeric channel ID (e.g. 'C0123456789'), not the #channel-name
JSON passthrough from provider API—
slack_join_channel
Join a public Slack channel so the bot can read history and post messages. Private channels require an invite from a member and cannot be joined with this tool.
  • channelId: string — Alphanumeric public channel ID to join (e.g. 'C0123456789')
JSON passthrough from provider API—
slack_send_message
Post a new message to a Slack channel or thread. Use to notify users, deliver reports, or respond to conversations. Returns the timestamp ID ('ts') of the posted message.
  • channel: string — Target channel ID, group DM ID, or user direct message ID (e.g. 'C0123456789')
  • text: string — Primary message body text supporting Slack mrkdwn formatting
  • threadTs?: string — Parent message timestamp ID (e.g. '1620000000.000100') to reply in a sub-thread instead of main channel
  • blocks?: array<object> — Optional array of Slack Block Kit UI blocks for rich layout (sections, buttons, dividers)
JSON passthrough from provider API—
slack_update_message
Edit the text content of a message previously posted by the bot. Use to update ongoing progress indicators, mark tasks completed, or correct typos.
  • channel: string — Channel ID containing the message to be updated (e.g. 'C0123456789')
  • ts: string — Timestamp identifier ('ts') of the exact message to edit (e.g. '1620000000.000100')
  • text: string — Replacement message text supporting Slack mrkdwn formatting
JSON passthrough from provider API—
slack_delete_message
Permanently delete a message from a Slack channel or thread. Note: Can only delete messages authored by the bot unless the bot has workspace admin privileges.
  • channel: string — Channel ID containing the message to delete
  • ts: string — Timestamp identifier ('ts') of the message to delete
JSON passthrough from provider API—
slack_get_thread
Fetch all reply messages from a specific conversation thread in chronological order. Use to read multi-turn discussions under a parent message.
  • channel: string — Channel ID where the parent thread exists
  • threadTs: string — Timestamp ID of the parent message that started the thread
  • limit?: number — Maximum number of thread replies to return (default 50, max 100)
JSON passthrough from provider API—
slack_search_messages
Search across workspace messages matching keywords, filters (from:@user, in:#channel), and date ranges. Use to find prior discussions and decisions.
  • query: string — Search query syntax including keywords and filters (e.g. 'project deadline from:@sarah in:#eng')
  • sort?: string — Result ranking order: 'score' ranks by semantic and keyword relevance, 'timestamp' sorts in reverse chronological order
  • count?: number — Maximum number of matching search results to return (min 1, max 100, default 20)
JSON passthrough from provider API—
slack_get_channel_history
Read recent messages from a channel's main timeline in reverse chronological order. Does not expand thread replies; use slack_get_thread for full threads.
  • channel: string — Channel ID to read message history from
  • limit?: number — Maximum number of recent messages to return (min 1, max 100, default 20)
  • oldest?: string — Unix epoch timestamp boundary; messages sent before this timestamp are excluded
  • latest?: string — Unix epoch timestamp boundary; messages sent after this timestamp are excluded
JSON passthrough from provider API—
slack_add_reaction
Add an emoji reaction to a specific message in a channel. Use to acknowledge notifications, cast votes, or add emoji feedback.
  • channel: string — Channel ID containing the message to react to (e.g. 'C0123456789')
  • timestamp: string — Timestamp identifier ('ts') of the message to receive the reaction
  • emoji: string — Standard emoji name string without colon delimiters (e.g. 'thumbsup' or 'eyes')
JSON passthrough from provider API—
slack_remove_reaction
Remove an emoji reaction previously added by the bot from a message.
  • channel: string — Channel ID containing the message from which to remove the reaction
  • timestamp: string — Timestamp identifier ('ts') of the message with the reaction
  • emoji: string — Standard emoji name string without colon delimiters to remove
JSON passthrough from provider API—
slack_list_users
List all user accounts in the Slack workspace including real names, display names, email addresses, and bot status.
  • limit?: number — Maximum number of user profile objects to return (min 1, max 200, default 100)
JSON passthrough from provider API—
slack_get_user_info
Fetch detailed profile information for a Slack user including real name, timezone, email, status text, and admin roles.
  • userId: string — Alphanumeric Slack user ID (e.g. 'U0123456789'), not the @mention username
JSON passthrough from provider APIparam: `userId` (caller-asserted)
slack_get_user_presence
Check whether a Slack user is currently actively logged in ('active') or away/offline ('away').
  • userId: string — Alphanumeric Slack user ID to check (e.g. 'U0123456789')
JSON passthrough from provider APIparam: `userId` (caller-asserted)
slack_upload_file
Upload a file, log snippet, or document directly to one or more Slack channels. Use to share generated reports, code snippets, or export files.
  • channels: string — Comma-separated list of channel IDs where the file will be shared
  • content: string — Raw textual content or code string of the file to be uploaded
  • filename: string — File name including file extension (e.g. 'test_output.json' or 'report.csv')
  • filetype?: string — Slack syntax highlight filetype identifier (e.g. 'json', 'python', 'javascript', 'csv')
  • title?: string — Human-readable title displayed above the file preview in the Slack client
JSON passthrough from provider API—
Microsoft Teams teams · 10 tools · Per-user OAuth / credentials

Microsoft Teams chat, channels, and meetings via delegated Microsoft Graph (/me). People-lane user OAuth — not application permissions.

Transport: cloud + local (stdio) · Required credentials: MICROSOFT_TENANT_ID, MICROSOFT_CLIENT_ID, MICROSOFT_CLIENT_SECRET, MICROSOFT_ACCESS_TOKEN

ToolParametersReturnsUser Scoping
teams_list_teams
List all teams the current user has joined
noneJSON passthrough from provider API—
teams_list_channels
List all standard and private channels within a specific Microsoft Teams team.
  • teamId: string — Unique Microsoft Teams group ID GUID (e.g. '00000000-0000-0000-0000-000000000000')
JSON passthrough from provider API—
teams_send_message
Send a message to a team channel
  • teamId: string — Unique Microsoft Teams group ID GUID (e.g. '00000000-0000-0000-0000-000000000000')
  • channelId: string — Unique channel ID within the team (e.g. '19:...@thread.tacv2')
  • content: string — Message content (HTML supported)
  • contentType?: string — Message body formatting type: 'text' for plain text or 'html' for rich HTML markup
JSON passthrough from provider API—
teams_reply_message
Reply to a message in a team channel
  • teamId: string — Unique Microsoft Teams group ID GUID (e.g. '00000000-0000-0000-0000-000000000000')
  • channelId: string — Unique channel ID within the team (e.g. '19:...@thread.tacv2')
  • messageId: string — Parent message ID
  • content: string — Reply content (HTML supported)
  • contentType?: string — Message body formatting type: 'text' for plain text or 'html' for rich HTML markup
JSON passthrough from provider API—
teams_list_messages
List messages in a team channel
  • teamId: string — Unique Microsoft Teams group ID GUID (e.g. '00000000-0000-0000-0000-000000000000')
  • channelId: string — Unique channel ID within the team (e.g. '19:...@thread.tacv2')
  • top?: number — Number of messages to return
JSON passthrough from provider API—
teams_list_chats
List all chats for the current user
noneJSON passthrough from provider API—
teams_send_chat
Send a message in a 1:1 or group chat
  • chatId: string — Unique Microsoft Teams 1:1 or group chat conversation ID
  • content: string — Message content (HTML supported)
  • contentType?: string — Message body formatting type: 'text' for plain text or 'html' for rich HTML markup
JSON passthrough from provider API—
teams_get_chat_messages
Get messages from a 1:1 or group chat
  • chatId: string — Unique Microsoft Teams 1:1 or group chat conversation ID
  • top?: number — Number of messages to return
JSON passthrough from provider API—
teams_create_meeting
Schedule an online Microsoft Teams meeting with subject, start/end timestamps, and join URL.
  • subject: string — Meeting subject
  • startDateTime: string — Start time in ISO 8601 format (e.g. 2026-03-20T10:00:00Z)
  • endDateTime: string — End time in ISO 8601 format
  • attendees?: array<string> — Attendee email addresses
JSON passthrough from provider APIparam: `attendees` (caller-asserted)
teams_list_meetings
Look up an online meeting by exact joinWebUrl or joinMeetingId. Graph has no unfiltered or date-window list of /me/onlineMeetings.
  • joinWebUrl?: string — Exact meeting join URL (JoinWebUrl eq filter)
  • joinMeetingId?: string — Teams join meeting ID (joinMeetingIdSettings/joinMeetingId eq filter)
JSON passthrough from provider API—
Zoom zoom · 10 tools · Organization credentials

Zoom integration for scheduling meetings, managing recordings, fetching transcripts, and participant management via Zoom API.

Transport: cloud + local (stdio) · Required credentials: ZOOM_CLIENT_ID, ZOOM_CLIENT_SECRET, ZOOM_ACCOUNT_ID

ToolParametersReturnsUser Scoping
zoom_list_meetings
List meetings for the authenticated user
  • type?: string — Meeting type filter (default: scheduled)
  • pageSize?: number — Results per page (max 300)
  • nextPageToken?: string — Pagination token
JSON passthrough from provider API—
zoom_get_meeting
Get detailed information about a specific meeting
  • meetingId: number — The meeting ID
JSON passthrough from provider API—
zoom_create_meeting
Create a new Zoom meeting
  • topic: string — Meeting topic/title
  • type?: number — 1=instant, 2=scheduled, 3=recurring no fixed time, 8=recurring fixed time
  • startTime?: string — Start time in ISO 8601 (for scheduled meetings)
  • duration?: number — Duration in minutes
  • timezone?: string — Timezone (e.g. America/New_York)
  • agenda?: string — Meeting agenda/description
  • settings?: object — Meeting settings (host_video, participant_video, join_before_host, mute_upon_entry, waiting_room, auto_recording)
JSON passthrough from provider API—
zoom_update_meeting
Update an existing meeting
  • meetingId: number — The meeting ID
  • topic?: string — Updated meeting topic
  • type?: number — Meeting type
  • startTime?: string — Updated start time in ISO 8601
  • duration?: number — Updated duration in minutes
  • timezone?: string — Timezone (e.g. America/New_York)
  • agenda?: string — Updated agenda
  • settings?: object — Updated meeting settings
JSON passthrough from provider API—
zoom_delete_meeting
Permanently delete a scheduled Zoom meeting by meeting ID and notify participants.
  • meetingId: number — The meeting ID to delete
JSON passthrough from provider API—
zoom_list_participants
List participants of a past meeting
  • meetingId: string — The past meeting ID or UUID
  • pageSize?: number — Results per page (max 300)
  • nextPageToken?: string — Pagination token
JSON passthrough from provider API—
zoom_get_recording
Get recording files for a specific meeting
  • meetingId: string — The meeting ID or UUID
JSON passthrough from provider API—
zoom_list_recordings
List cloud recordings for the authenticated user within a date range
  • from: string — Start date (YYYY-MM-DD)
  • to: string — End date (YYYY-MM-DD)
  • pageSize?: number — Results per page (max 300)
  • nextPageToken?: string — Pagination token
JSON passthrough from provider API—
zoom_create_webinar
Schedule a new Zoom webinar with registration settings, panelist controls, and topic.
  • topic: string — Webinar topic/title
  • type?: number — 5=webinar, 6=recurring no fixed time, 9=recurring fixed time
  • startTime?: string — Start time in ISO 8601
  • duration?: number — Duration in minutes
  • timezone?: string — Timezone (e.g. America/New_York)
  • agenda?: string — Webinar agenda/description
  • settings?: object — Webinar settings
JSON passthrough from provider API—
zoom_list_webinars
List webinars for the authenticated user
  • pageSize?: number — Results per page (max 300)
  • nextPageToken?: string — Pagination token
JSON passthrough from provider API—

data (1)

Nunc Memory nunc-memory · 6 tools · Built-in (no external credentials)

Persistent memory and knowledge graph for AI conversations. Stores facts, preferences, and context across sessions with hybrid vector + keyword search.

Transport: cloud + local (stdio)

ToolParametersReturnsUser Scoping
memory_recall
Search and recall memories relevant to the current context
  • query: string — Search query to find relevant memories
  • userId: string — User ID for tenant-scoped memory
  • workspaceId?: string — Optional workspace ID for project-scoped memory
  • scopeId?: string — Optional explicit scope key ('company', 'workspace:<id>', 'user:<id>')
  • limit?: number — Maximum number of memories to return
JSON passthrough from provider APIparam: `userId` (caller-asserted)
memory_store
Store a new memory or conversation snippet
  • content: string — Memory content to store
  • userId: string — User ID for tenant-scoped memory
  • workspaceId?: string — Optional workspace ID to scope memory to project
  • scopeId?: string — Optional explicit scope key ('company', 'workspace:<id>', 'user:<id>')
  • metadata?: object — Optional metadata (tags, categories, etc.)
JSON passthrough from provider APIparam: `userId` (caller-asserted)
memory_search
Semantic search across all memories
  • query: string — Semantic search query
  • userId: string — User ID for tenant-scoped memory
  • workspaceId?: string — Optional workspace ID for project-scoped memory
  • scopeId?: string — Optional explicit scope key ('company', 'workspace:<id>', 'user:<id>')
  • filters?: object — Optional filters (date range, categories, etc.)
JSON passthrough from provider APIparam: `userId` (caller-asserted)
memory_delete
Delete a specific memory by ID
  • memoryId: string — Memory ID to delete
  • userId: string — User ID for authorization
JSON passthrough from provider APIparam: `userId` (caller-asserted)
memory_forget
Alias for memory_delete. Delete a memory by its ID
  • memoryId: string — Memory ID to delete
  • userId: string — User ID for authorization
JSON passthrough from provider APIparam: `userId` (caller-asserted)
memory_list
List stored memories with optional filtering and pagination
  • userId?: string — Optional User ID to scope list
  • node_type?: string — Optional node type filter (e.g. 'memory', 'doc')
  • category?: string — Optional category filter (e.g. 'conversation', 'preference')
  • limit?: number — Maximum number of memories to return (default 20)
  • offset?: number — Pagination offset (default 0)
JSON passthrough from provider APIparam: `userId` (caller-asserted)

design (3)

Adobe Creative Cloud adobe · 7 tools · Organization credentials

Adobe Creative Cloud integration for Photoshop, Illustrator, and InDesign automation. Supports asset management, batch processing, and template rendering.

Transport: cloud + local (stdio) · Required credentials: ADOBE_ACCESS_TOKEN, ADOBE_CLIENT_ID, ADOBE_ORG_ID

ToolParametersReturnsUser Scoping
adobe_remove_background
Remove the background from an image using Adobe Sensei
  • inputUrl: string — URL of the input image
  • outputUrl: string — URL to store the output image
JSON passthrough from provider API—
adobe_smart_crop
Auto-crop an image to its subject using Adobe Sensei
  • inputUrl: string — URL of the input image
  • outputUrl: string — URL to store the cropped image
  • width: number — Target width in pixels
  • height: number — Target height in pixels
JSON passthrough from provider API—
adobe_create_rendition
Convert a PSD file to PNG or JPG rendition
  • inputUrl: string — URL of the PSD file
  • outputUrl: string — URL to store the rendition
  • format: string — Output format
  • width?: number — Output width in pixels
  • quality?: number — JPEG quality (1-100, only for jpg)
JSON passthrough from provider API—
adobe_replace_smart_object
Replace the content of a smart object layer in a PSD
  • inputUrl: string — URL of the PSD file
  • outputUrl: string — URL to store the modified PSD
  • layerName: string — Name of the smart object layer
  • contentUrl: string — URL of the replacement content
JSON passthrough from provider API—
adobe_apply_preset
Apply a Lightroom preset to an image
  • inputUrl: string — URL of the input image
  • outputUrl: string — URL to store the processed image
  • presetId: string — Lightroom preset identifier
JSON passthrough from provider API—
adobe_generate_image
Generate an image from a text prompt using Adobe Firefly
  • prompt: string — Text description of the image to generate
  • outputUrl: string — URL to store the generated image
  • width?: number — Image width in pixels
  • height?: number — Image height in pixels
JSON passthrough from provider API—
adobe_expand_image
Expand an image canvas with generative fill using Adobe Firefly
  • inputUrl: string — URL of the input image
  • outputUrl: string — URL to store the expanded image
  • width: number — Target canvas width
  • height: number — Target canvas height
JSON passthrough from provider API—
Canva canva · 9 tools · Organization credentials

Canva integration for creating designs, managing brand kits, exporting assets, and accessing templates via Canva Connect API.

Transport: cloud + local (stdio) · Required credentials: CANVA_ACCESS_TOKEN

ToolParametersReturnsUser Scoping
canva_list_designs
List designs for the authenticated user
noneJSON passthrough from provider API—
canva_get_design
Get details of a specific design
  • designId: string — Canva design ID
JSON passthrough from provider API—
canva_create_design
Create a new Canva design
  • title: string — Design title
  • designType: string — Type of design to create
  • width?: number — Optional custom width in pixels (requires height)
  • height?: number — Optional custom height in pixels (requires width)
JSON passthrough from provider API—
canva_list_folders
List items in a Canva folder (defaults to root folder)
  • folderId?: string — Canva folder ID (defaults to 'root')
JSON passthrough from provider API—
canva_upload_asset
Upload an asset (image/file) to Canva
  • name: string — Asset name
  • fileData: string — Base64-encoded file data
JSON passthrough from provider API—
canva_list_templates
List brand and design templates accessible in the Canva workspace. Returns template IDs, titles, and preview URLs.
noneJSON passthrough from provider API—
canva_export_design
Export a design to PDF, PNG, or JPG
  • designId: string — Canva design ID
  • format: string — Export format
  • quality?: string — Export quality
JSON passthrough from provider API—
canva_get_export_status
Check the status of a design export
  • exportId: string — Export job ID
JSON passthrough from provider API—
canva_autofill
Create a design by autofilling a brand template with data
  • templateId: string — Brand template ID to fill
  • data: object — Key-value pairs to fill template placeholders
JSON passthrough from provider API—
Figma figma · 12 tools · Organization credentials

Figma design integration for reading files, extracting components, inspecting styles, and exporting assets. Supports design tokens and variable extraction.

Transport: cloud + local (stdio) · Required credentials: FIGMA_ACCESS_TOKEN

ToolParametersReturnsUser Scoping
figma_list_files
List files in a Figma project by project ID
  • projectId: string — Figma project ID
JSON passthrough from provider API—
figma_get_file
Get the full file tree for a Figma file
  • fileKey: string — Figma file key
JSON passthrough from provider API—
figma_get_file_nodes
Get specific nodes from a Figma file by node IDs
  • fileKey: string — Figma file key
  • nodeIds: string — Comma-separated node IDs
JSON passthrough from provider API—
figma_get_comments
Get all comments on a Figma file
  • fileKey: string — Figma file key
JSON passthrough from provider API—
figma_post_comment
Post a comment on a Figma file
  • fileKey: string — Figma file key
  • message: string — Comment text
  • x?: number — X position for the comment pin
  • y?: number — Y position for the comment pin
JSON passthrough from provider API—
figma_export_nodes
Export nodes as images (PNG, SVG, or PDF)
  • fileKey: string — Figma file key
  • nodeIds: string — Comma-separated node IDs to export
  • format?: string — Export format
  • scale?: number — Export scale (1-4)
JSON passthrough from provider API—
figma_get_styles
Get design tokens (colors, text styles, effects) from a Figma file
  • fileKey: string — Figma file key
JSON passthrough from provider API—
figma_get_components
List reusable components in a Figma file
  • fileKey: string — Figma file key
JSON passthrough from provider API—
figma_get_component_set
Get component sets (variant groups) from a Figma file
  • fileKey: string — Figma file key
JSON passthrough from provider API—
figma_list_projects
List projects within a Figma team. Use to discover project IDs before querying files.
  • teamId: string — Figma team ID
JSON passthrough from provider API—
figma_list_project_files
List all design and FigJam files within a specific Figma project. Returns file keys and names.
  • projectId: string — Figma project ID
JSON passthrough from provider API—
figma_get_variables
Get local design variables/tokens from a Figma file
  • fileKey: string — Figma file key
JSON passthrough from provider API—

development (1)

GitHub github · 39 tools · Organization credentials

Unified Git and GitHub integration for local repository operations (status, diff, log, branch, commit, push, worktree) and GitHub API capabilities (issues, pull requests, code reviews, workflows).

Transport: cloud + local (stdio) · Required credentials: GIT_REPO_PATH, GITHUB_PERSONAL_ACCESS_TOKEN, CLIENT_ID, CLIENT_SECRET

ToolParametersReturnsUser Scoping
github_list_repos
List repositories accessible to the authenticated user. Use this to discover repository names and permissions across personal and organizational workspaces. Returns an array of repository summaries.
  • org?: string — Optional organization name to list repos for; omit to list repositories owned by the authenticated user
  • sort?: string — Sort order: 'created' orders by creation date, 'updated' by latest modification, 'pushed' by latest commit push, 'full_name' alphabetically
  • perPage?: number — Number of repository results per page (min 1, max 100, default 30)
JSON passthrough from provider API—
github_get_repo
Get comprehensive metadata for a specific repository including default branch, star count, open issues count, license, and repository capabilities. Use before opening PRs or issues.
  • owner: string — GitHub organization or user account that owns the repository (e.g. 'facebook' or 'octocat')
  • repo: string — Repository name without the owner prefix (e.g. 'react')
JSON passthrough from provider APIparam: `owner` (caller-asserted)
github_list_issues
List issues for a repository. Use to query open bugs, tasks, or feature requests. Note: GitHub's issues API also returns pull requests unless filtered; check the 'pull_request' key in results.
  • owner: string — GitHub organization or user account that owns the repository (e.g. 'facebook' or 'octocat')
  • repo: string — Repository name without the owner prefix (e.g. 'react')
  • state?: string — Filter issues by lifecycle state: 'open' returns active unresolved issues, 'closed' returns resolved/dismissed issues, 'all' returns both
  • labels?: string — Comma-separated list of label names (e.g. 'bug,ui') to restrict results; issues must match all specified labels
  • assignee?: string — Filter issues assigned to a specific GitHub username; pass 'none' for unassigned issues or '*' for any assigned issue
  • sort?: string — Sort criteria: 'created' orders by issue creation date, 'updated' orders by latest comment or modification, 'comments' orders by discussion volume
  • perPage?: number — Number of issue records to return per page (min 1, max 100, default 30)
JSON passthrough from provider APIparam: `owner`, `assignee` (caller-asserted)
github_create_issue
Open a new issue in a repository. Use to file bug reports, track tasks, or propose improvements. Returns the newly created issue object with its URL and number.
  • owner: string — GitHub organization or user account that owns the repository (e.g. 'facebook' or 'octocat')
  • repo: string — Repository name without the owner prefix (e.g. 'react')
  • title: string — Short, descriptive summary line for the issue (displayed in lists and email notifications)
  • body?: string — Detailed description of the issue supporting GitHub Flavored Markdown, code fences, and @mentions
  • labels?: array<string> — Array of existing label names (e.g. ['bug', 'priority:high']) to attach upon creation
  • assignees?: array<string> — Array of GitHub usernames to assign to the issue; users must have triage or write permissions on the repository
JSON passthrough from provider APIparam: `owner`, `assignees` (caller-asserted)
github_update_issue
Modify an existing issue's title, body, status, or labels. Use to close issues, re-assign work, or update problem descriptions. Returns updated issue metadata.
  • owner: string — GitHub organization or user account that owns the repository (e.g. 'facebook' or 'octocat')
  • repo: string — Repository name without the owner prefix (e.g. 'react')
  • issueNumber: number — The 1-based sequential issue number (e.g. 42 for issue #42), distinct from internal database IDs
  • title?: string — New title string replacing the existing issue title
  • body?: string — New markdown body string replacing the full existing issue description
  • state?: string — State transition: 'open' re-opens a closed issue, 'closed' marks the issue as resolved
  • labels?: array<string> — Replacement array of label names; replaces all existing labels on the issue
JSON passthrough from provider APIparam: `owner` (caller-asserted)
github_add_comment
Post a markdown comment to an issue or pull request discussion thread. Use to provide feedback, report test results, or ask clarifying questions.
  • owner: string — GitHub organization or user account that owns the repository (e.g. 'facebook' or 'octocat')
  • repo: string — Repository name without the owner prefix (e.g. 'react')
  • issueNumber: number — The 1-based sequential issue number (e.g. 42 for issue #42), distinct from internal database IDs
  • body: string — Comment message text supporting Markdown, checklists, tables, and @mentions
JSON passthrough from provider APIparam: `owner` (caller-asserted)
github_list_prs
List pull requests in a repository. Use to review active code reviews, check CI readiness, or find branch integration status.
  • owner: string — GitHub organization or user account that owns the repository (e.g. 'facebook' or 'octocat')
  • repo: string — Repository name without the owner prefix (e.g. 'react')
  • state?: string — Filter PRs by lifecycle state: 'open' returns unmerged active PRs, 'closed' returns closed without merge, 'all' returns open, closed, and merged
  • head?: string — Filter by head branch or fork (format: 'owner:branch' or 'branch' within same repo)
  • base?: string — Filter by target integration branch (e.g. 'main', 'master', or 'develop')
  • sort?: string — Sort attribute: 'created' orders by opening date, 'updated' by latest review or commit, 'popularity' by comments/reactions
  • perPage?: number — Number of pull requests to return per page (min 1, max 100, default 30)
JSON passthrough from provider APIparam: `owner` (caller-asserted)
github_create_pr
Open a new pull request from a head branch into a base branch. Use to submit code changes for peer review. Returns the created PR with URL and PR number.
  • owner: string — GitHub organization or user account that owns the repository (e.g. 'facebook' or 'octocat')
  • repo: string — Repository name without the owner prefix (e.g. 'react')
  • title: string — Pull request title summarizing the proposed changes
  • body?: string — Markdown description detailing motivation, changes, and issue links (e.g. 'Closes #12')
  • head: string — The branch containing the new commits to be merged (e.g. 'feature/auth')
  • base: string — The target branch that will receive the merged commits (typically 'main')
  • draft?: boolean — When true, creates the pull request as a draft to prevent accidental early merging and suppress review requests
JSON passthrough from provider APIparam: `owner` (caller-asserted)
github_get_pr
Get full details of a specific pull request including mergeability status, diff stats, base/head commit SHAs, and review state.
  • owner: string — GitHub organization or user account that owns the repository (e.g. 'facebook' or 'octocat')
  • repo: string — Repository name without the owner prefix (e.g. 'react')
  • prNumber: number — The 1-based sequential pull request number (e.g. 101 for PR #101)
JSON passthrough from provider APIparam: `owner` (caller-asserted)
github_merge_pr
Merge an approved pull request into the base branch. Use when reviews and automated status checks have passed. Returns merge commit details.
  • owner: string — GitHub organization or user account that owns the repository (e.g. 'facebook' or 'octocat')
  • repo: string — Repository name without the owner prefix (e.g. 'react')
  • prNumber: number — The 1-based sequential pull request number (e.g. 101 for PR #101)
  • commitTitle?: string — Title line for the merge commit
  • commitMessage?: string — Detailed description body for the merge commit
  • mergeMethod?: string — Merge strategy: 'merge' creates a merge commit preserving all commits, 'squash' combines all commits into a single commit on base, 'rebase' replays commits linearly onto base
JSON passthrough from provider APIparam: `owner` (caller-asserted)
github_list_pr_reviews
List all review submissions on a pull request, showing reviewers, approval states (APPROVED, CHANGES_REQUESTED, COMMENTED), and submitted timestamps.
  • owner: string — GitHub organization or user account that owns the repository (e.g. 'facebook' or 'octocat')
  • repo: string — Repository name without the owner prefix (e.g. 'react')
  • prNumber: number — The 1-based sequential pull request number (e.g. 101 for PR #101)
JSON passthrough from provider APIparam: `owner` (caller-asserted)
github_create_pr_review
Submit a formal code review on a pull request with an optional decision. Use to approve changes or request modifications before merging.
  • owner: string — GitHub organization or user account that owns the repository (e.g. 'facebook' or 'octocat')
  • repo: string — Repository name without the owner prefix (e.g. 'react')
  • prNumber: number — The 1-based sequential pull request number (e.g. 101 for PR #101)
  • body?: string — Summary review comment text explaining the rationale, requested changes, or approval notes
  • event: string — Review decision action: 'APPROVE' submits approval allowing merge, 'REQUEST_CHANGES' blocks merging until resolved, 'COMMENT' submits feedback without an explicit decision
JSON passthrough from provider APIparam: `owner` (caller-asserted)
github_list_workflow_runs
List GitHub Actions CI/CD workflow executions for a repository. Use to inspect build, test, and deployment statuses.
  • owner: string — GitHub organization or user account that owns the repository (e.g. 'facebook' or 'octocat')
  • repo: string — Repository name without the owner prefix (e.g. 'react')
  • workflowId?: string — Optional workflow filename (e.g. 'deploy.yml') or numerical workflow ID to filter runs by
  • status?: string — Execution status filter: 'completed' finished execution, 'in_progress' currently executing, 'queued' awaiting an available runner, 'failure' ended with errors
  • perPage?: number — Number of workflow run records to return (min 1, max 100, default 10)
JSON passthrough from provider APIparam: `owner` (caller-asserted)
github_trigger_workflow
Trigger a workflow_dispatch event to run a GitHub Actions workflow manually. Returns HTTP 204 on successful dispatch.
  • owner: string — GitHub organization or user account that owns the repository (e.g. 'facebook' or 'octocat')
  • repo: string — Repository name without the owner prefix (e.g. 'react')
  • workflowId: string — Workflow filename (e.g. 'deploy.yml') or numerical workflow ID
  • ref: string — Git branch or tag reference that the workflow will execute against (e.g. 'main' or 'v1.2.0')
  • inputs?: object — Key-value input parameters defined in the workflow's workflow_dispatch schema
JSON passthrough from provider APIparam: `owner` (caller-asserted)
github_get_file_contents
Read raw or decoded file contents from a repository at a specific branch or commit. Use to inspect configuration files or source code without cloning.
  • owner: string — GitHub organization or user account that owns the repository (e.g. 'facebook' or 'octocat')
  • repo: string — Repository name without the owner prefix (e.g. 'react')
  • path: string — Path to the file relative to repository root (e.g. 'packages/core/src/index.ts')
  • ref?: string — Commit SHA, branch name, or tag to read from; defaults to the repository's default branch
JSON passthrough from provider APIparam: `owner` (caller-asserted)
github_search_code
Search code across GitHub repositories using GitHub search qualifiers (e.g. 'filename:package.json language:typescript'). Returns matching code snippets and locations.
  • query: string — Search query syntax including keywords and qualifiers (e.g. 'auth repo:owner/repo path:src/')
  • sort?: string — Optional sort field ('indexed' sorts by how recently a file was indexed by GitHub)
  • perPage?: number — Number of search results per page (min 1, max 100, default 30)
JSON passthrough from provider API—
git_status
Get working tree status showing modified, staged, untracked, and deleted files. Use before committing or branching.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
JSON passthrough from provider API—
git_log
Retrieve chronological commit history for the current branch or a specified revision. Use to review recent changes or find commit SHAs.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • maxCount?: number — Maximum number of commit history entries to return (default 20, max 100)
  • branch?: string — Branch, tag, or commit reference to inspect (defaults to current active HEAD)
  • oneline?: boolean — When true, formats each commit as a single line with abbreviated SHA and commit subject
JSON passthrough from provider API—
git_diff
Show unified line diffs of unstaged changes, staged changes, or differences between two commits/branches.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • staged?: boolean — When true, shows changes staged in the index ready for commit instead of unstaged working tree diffs
  • file?: string — Relative file path to restrict the diff output to a single file
  • commitA?: string — Base commit SHA, branch name, or tag for two-point comparison (e.g. 'main')
  • commitB?: string — Target commit SHA or branch to compare against commitA (defaults to working tree if omitted)
JSON passthrough from provider API—
git_blame
Annotate lines of a file with commit SHAs, authors, and dates to determine who authored each line. Use to trace regressions.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • file: string — Relative path to the file to inspect with blame annotations
  • lineStart?: number — 1-based starting line number to bound blame output
  • lineEnd?: number — 1-based ending line number to bound blame output
JSON passthrough from provider API—
git_add
Stage file changes into the index for the next commit. Use to select specific files or all changes.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • files: array<string> — List of file paths to stage; pass ['.'] to stage all modified and untracked files
JSON passthrough from provider API—
git_commit
Create a new commit recording staged changes with a descriptive message.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • message: string — Commit message describing the changes following Conventional Commits format
  • amend?: boolean — When true, amends and replaces the most recent commit on the current branch
JSON passthrough from provider API—
git_reset
Unstage files or reset working tree. Hard resets are destructive and require explicit confirm: true.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • files?: array<string> — Specific file paths to remove from the staging index (omit to reset all staged files)
  • hard?: boolean — When true, performs a hard reset discarding both staged and unstaged working tree modifications (DESTRUCTIVE)
  • confirm?: boolean — Safety confirmation flag; must be set to true when hard is true to confirm discarding uncommitted work
JSON passthrough from provider API—
git_branch_list
List local and optional remote-tracking branches in the repository.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • remote?: boolean — When true, includes remote tracking branches (origin/*) in addition to local branches
JSON passthrough from provider API—
git_branch_create
Create a new Git branch starting from HEAD or a specified commit ref without switching to it.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • name: string — Name for the new Git branch (e.g. 'feat/oauth-login')
  • startPoint?: string — Commit SHA, branch name, or tag to branch off from (defaults to current HEAD)
JSON passthrough from provider API—
git_checkout
Switch the active working branch or restore files from the index. Supports creating and switching via create flag.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • branch: string — Target branch name to switch to
  • create?: boolean — When true, creates the branch if it does not already exist (equivalent to 'git checkout -b')
JSON passthrough from provider API—
git_merge
Merge changes from another branch into the currently checked-out branch.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • branch: string — Name of the branch to merge into the active branch
  • noFf?: boolean — When true, always creates a merge commit even if the merge could resolve as a fast-forward
JSON passthrough from provider API—
git_push
Push local branch commits to a remote Git repository. Force push is destructive and requires confirm: true.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • remote?: string — Remote repository name to push to (defaults to 'origin')
  • branch?: string — Branch name to push (defaults to current tracking branch)
  • force?: boolean — When true, overwrites remote history with force-with-lease (DESTRUCTIVE — requires confirm: true)
  • confirm?: boolean — Safety confirmation flag; must be set to true when force is true
JSON passthrough from provider API—
git_pull
Fetch and integrate changes from a remote repository into the active branch via merge or rebase.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • remote?: string — Remote name to pull from (defaults to 'origin')
  • branch?: string — Remote branch name to pull (defaults to current tracking branch)
  • rebase?: boolean — When true, replays local commits on top of incoming upstream commits instead of creating a merge commit
JSON passthrough from provider API—
git_remote_list
List all configured remote repositories with their fetch and push URLs.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
JSON passthrough from provider API—
git_stash
Save uncommitted working tree and staged changes into a temporary stash entry, resetting working tree to HEAD.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • message?: string — Optional description message to label the stash entry
JSON passthrough from provider API—
git_stash_pop
Apply the most recent stash entry back into the working tree and remove it from the stash stack.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • index?: number — 0-based stash index to pop (defaults to 0 for most recent)
JSON passthrough from provider API—
git_worktree_list
List all active linked git worktrees with their paths, HEAD commit SHAs, and active branches.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
JSON passthrough from provider API—
git_worktree_add
Create and checkout a new linked worktree in an isolated directory.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • path: string — Filesystem path where the new worktree will be created
  • branch?: string — Branch to checkout in the new worktree
  • newBranch?: boolean — When true, creates a new branch with the specified name
  • base?: string — Base commit or branch to branch off from if newBranch is true
JSON passthrough from provider API—
git_worktree_remove
Remove a linked worktree and clean up administrative files. Forced removal requires confirm: true.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • path: string — Path of the linked worktree to remove
  • force?: boolean — When true, forces removal even if the worktree contains uncommitted changes (requires confirm: true)
  • confirm?: boolean — Safety confirmation flag; must be set to true when force is true
JSON passthrough from provider API—
git_cherry_pick
Apply the changes introduced by one or more existing commits onto the current branch.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • commits: array<string> — List of commit SHAs or refs to cherry-pick in sequential order
  • noCommit?: boolean — When true, applies the commit changes to working tree and index without creating commits
JSON passthrough from provider API—
git_rebase
Reapply commits on top of another base tip, or abort/continue an ongoing rebase.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • onto?: string — Target upstream branch or commit to rebase onto
  • abort?: boolean — When true, aborts the current rebase and resets to original HEAD
  • continue_rebase?: boolean — When true, continues rebase after resolving merge conflicts
JSON passthrough from provider API—
git_tag
List, create, or delete Git tags.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • action: string — Tag operation to perform: 'list', 'create', or 'delete'
  • name?: string — Tag name for create or delete actions
  • message?: string — Annotation message for annotated tag creation
  • commit?: string — Specific commit SHA to tag (defaults to current HEAD)
JSON passthrough from provider API—
git_pr_prepare
Prepare a branch, stage files, commit changes, and push upstream to origin ready for opening a pull request.
  • repoPath?: string — Absolute or workspace-relative path to the target Git repository root
  • branch: string — The head branch name to create/checkout and push
  • title: string — Commit title message for the changes
  • base?: string — Base branch (defaults to 'main')
  • files?: array<string> — Optional list of specific files to stage (omit to keep existing staged changes)
JSON passthrough from provider API—

engineering (1)

3D Modeling modeling-3d · 12 tools · Built-in (no external credentials)

General-purpose 3D modeling integration for mesh generation, STL/OBJ processing, parametric design, and 3D preview rendering.

Transport: cloud + local (stdio)

ToolParametersReturnsUser Scoping
model3d_create_box
Generate a 3D box (rectangular prism) mesh. Returns vertex, normal, and face data for the Nunc 3D CAD renderer.
  • width?: number — Dimension along the X axis in world units
  • height?: number — Dimension along the Y axis in world units
  • depth?: number — Dimension along the Z axis in world units
  • position?: object — Center placement coordinates in 3D world space
  • rotation?: object — Euler rotation angles in degrees around the X, Y, and Z axes
  • color?: string — Hexadecimal surface material color (e.g. '#2563eb' for blue, '#10b981' for green)
JSON passthrough from provider API—
model3d_create_sphere
Generate a 3D sphere mesh centered at a position. Returns tessellated vertex and normal buffers for rendering.
  • radius?: number — Distance from center to surface in world units
  • segments?: number — Number of latitude and longitude subdivision segments (higher values produce smoother curvature at higher polygon cost)
  • position?: object — Center position in 3D world space
  • color?: string — Hexadecimal surface color code (e.g. '#e11d48')
JSON passthrough from provider API—
model3d_create_cylinder
Generate a 3D cylinder or truncated cone mesh. Set radiusTop to 0 to create a pointed cone.
  • radiusTop?: number — Radius of the circular top cap in world units (pass 0 to produce a cone)
  • radiusBottom?: number — Radius of the circular base cap in world units
  • height?: number — Vertical cylinder height along the Y axis
  • position?: object — Base center position in 3D world space
  • color?: string — Hexadecimal surface material color (e.g. '#0284c7')
JSON passthrough from provider API—
model3d_create_plane
Generate a flat rectangular 2D plane mesh in 3D space. Useful for floor grounds, backdrops, or sectional cuts.
  • width?: number — Horizontal width in world units
  • height?: number — Vertical height in world units
  • position?: object — Center placement position in 3D world space
  • rotation?: object — Euler rotation angles in degrees around X, Y, and Z
  • color?: string — Hexadecimal surface color code
JSON passthrough from provider API—
model3d_create_torus
Generate a 3D torus (donut ring) geometry mesh. Returns full ring surface vertex and normal buffers.
  • radius?: number — Major radius: distance from the center of the torus to the center of the circular tube
  • tube?: number — Minor radius: thickness radius of the circular tube itself
  • position?: object — Center position in 3D world space
  • color?: string — Hexadecimal surface material color
JSON passthrough from provider API—
model3d_boolean_operation
Perform constructive solid geometry (CSG) boolean operations (union, subtract, intersect) on two mesh geometries.
  • operationA: object — First input geometry mesh object (base solid)
  • operationB: object — Second input geometry mesh object (tool solid)
  • operation: string — CSG operation type: 'union' merges volumes, 'subtract' cuts tool out of base, 'intersect' retains shared volume
JSON passthrough from provider API—
model3d_transform
Apply translation, rotation, and scaling matrix transformations to a 3D geometry mesh.
  • geometry: object — Input geometry mesh object to transform
  • translate?: object — Position offset vector to translate the geometry along X, Y, Z
  • rotate?: object — Euler rotation angles in degrees to rotate around X, Y, Z axes
  • scale?: object — Dimensional scaling multipliers along X, Y, Z axes (1.0 = unchanged)
JSON passthrough from provider API—
model3d_extrude
Extrude a 2D planar polygon cross-section into a 3D solid mesh along the Z axis. Returns generated solid geometry.
  • points: array<object> — Ordered list of 2D polygon vertices forming the boundary profile (will automatically connect last point to first)
  • height?: number — Extrusion depth distance along the perpendicular Z axis
  • bevel?: number — Bevel radius in world units applied to outer boundary edges (pass 0 for crisp sharp corners)
JSON passthrough from provider API—
model3d_lathe
Generate a rotational 3D surface of revolution (lathe) by revolving a 2D curve 360 degrees around the vertical Y axis.
  • points: array<object> — Ordered 2D coordinate curve points defining the radial silhouette to revolve
  • segments?: number — Number of angular subdivision slices around the 360-degree rotation (higher values yield smoother round surfaces)
JSON passthrough from provider API—
model3d_export
Export composed 3D geometries to standard 3D CAD and visualization interchange formats (glTF, OBJ, or STL).
  • scene: array<object> — Array of geometry objects to include in the exported model file
  • format?: string — Output file format: 'gltf' for modern web CAD/WebGL scenes, 'obj' for universal mesh geometry, 'stl' for 3D printing slicing
JSON passthrough from provider API—
model3d_scene
Compose a complete 3D scene from multiple geometries, configuring view camera and illuminations for CAD viewport rendering.
  • objects?: array<object> — Array of geometry mesh definitions to place in the composed scene
  • camera?: object — Viewport camera position, orientation, and lens parameters
  • lights?: array<object> — Array of illumination light sources positioned throughout the scene
JSON passthrough from provider API—
model3d_parametric
Generate continuous 3D surface geometry by evaluating parametric formulas over a 2D (u,v) parameter domain. Supports predefined shapes or algebraic equations.
  • formula: string — Parametric surface formula: select a built-in preset ('sphere', 'torus', 'klein', 'mobius', 'helix') or provide algebraic equations: 'x=...; y=...; z=...'
  • uRange?: object — Evaluation domain interval [min, max] for the first parameter variable 'u'
  • vRange?: object — Evaluation domain interval [min, max] for the second parameter variable 'v'
  • uSegments?: number — Subdivision step resolution along the 'u' parameter curve (higher values produce smoother mesh)
  • vSegments?: number — Subdivision step resolution along the 'v' parameter curve (higher values produce smoother mesh)
  • color?: string — Hexadecimal surface material color code
JSON passthrough from provider API—

events (2)

Cvent cvent · 27 tools · Organization credentials

Cvent event management integration for creating events, managing registrations, tracking attendees, sending event emails, reading engagement metrics, and reading seating charts and assignments.

Transport: cloud + local (stdio) · Required credentials: CVENT_CLIENT_ID, CVENT_CLIENT_SECRET, CVENT_BASE_URL, CVENT_AUTH_URL, CVENT_SCOPE

27 tools across events / attendees / contacts / sessions / engagement metrics (Attendee Insights) / event emails / seating (read-only — Cvent's REST API has no diagram-authoring or venue-list endpoints, and cannot email arbitrary addresses). Attendees + sessions are GLOBAL resources reached via `filter` expressions (event.id / contact.email / status). Per-user scoping: `cvent_get_attendee_registrations` (exact-email, cross-event) + the `email` filter on `cvent_list_attendees`. `cvent_list_events` sorts on start (`start:DESC` = most recent). Returns are raw JSON passthrough of the Cvent REST response.

ToolParametersReturnsUser Scoping
cvent_list_events
List Cvent events. Use this for "most recent event", "upcoming events", or to get an event id before cvent_get_event. Most recent / latest-starting: sort=start:DESC and a small limit (1–5). Upcoming: status=Upcoming (Cvent eventStatus; Active is accepted as an alias), startDateAfter=<today ISO 8601>, sort=start:ASC. Recently edited: modifiedSince=<ISO 8601>. Next page: cursor from paging.nextToken. Cvent only sorts on start — without sort the page is not recency-ordered. Returns raw JSON { data, paging }.
Cvent only sorts on `start`. Use sort=start:DESC + limit=1 for the most recent event.
  • status?: string — Lifecycle eventStatus (Upcoming, Ongoing, Completed, Cancelled, Archived) or planning status (Pending, Accepted). Active is treated as Upcoming.
  • startDateAfter?: string — Filter events starting on/after this ISO datetime or YYYY-MM-DD date (e.g. 2026-07-01T00:00:00Z or 2026-07-01)
  • startDateBefore?: string — Filter events starting on/before this ISO datetime or YYYY-MM-DD date (e.g. 2026-07-31T23:59:59Z or 2026-07-31)
  • modifiedSince?: string — Filter events added or updated on/after this ISO date
  • sort?: string — Sort by event start time. start:DESC = most recent / latest-starting first (use with limit=1 for a single answer). start:ASC = soonest-starting first. Cvent only sorts on start.
  • cursor?: string — Cvent paging token from a prior page's paging.nextToken
  • limit?: number — Max results to return (default 25)
JSON passthrough from provider API—
cvent_get_event
Get full Cvent event details by ID. If you only have a name or need the latest/upcoming event, call cvent_list_events first (sort=start:DESC or start:ASC) and use an id from data[].
  • eventId: string — Cvent event ID
JSON passthrough from provider API—
cvent_rotate_token
Mint a fresh Cvent OAuth access token and cache it until expires_in (Cvent: 60 minutes). Call once at the start of a poll; data GETs reuse the cached bearer. Does not fetch events or attendees.
noneJSON passthrough from provider API—
cvent_create_event
Create a new event in Cvent (asynchronous — returns an eventCreationId and status; poll cvent_get_event_creation_status until status is Created, then the event ID is returned there)
  • title: string — Event title
  • startDate: string — Start date/time (ISO 8601)
  • endDate: string — End date/time (ISO 8601)
  • timezone: string — Event timezone (e.g. America/New_York)
  • eventType: string — Event type (e.g. Meeting, Conference, Seminar, TrainingSession, Webinar, TradeShow, Dinner, Forum)
  • plannerEmail: string — Email of the Cvent planner user who owns the event
  • format?: string — Event format: In-person (default), Virtual, or Hybrid
  • language?: string — IETF language tag for the event (default en-US)
  • description?: string — Event description
  • capacity?: number — Max attendees
JSON passthrough from provider API—
cvent_get_event_creation_status
Check the status of an asynchronous event creation (Processing | Created | Failed); returns the new event ID once Created
  • eventCreationId: string — Creation ID returned by cvent_create_event
JSON passthrough from provider API—
cvent_update_event
Update an existing event (read-modify-write: fetches the event, merges your updates, PUTs the full resource — Cvent has no partial update)
  • eventId: string — Cvent event ID
  • updates: object — Fields to update, using Cvent's field names (title, start, end, description, capacity, timezone, status, ...)
JSON passthrough from provider API—
cvent_list_attendees
List attendees globally or filtered by event ID, email, or status
`email` = exact identity filter (per-user scoping).
  • eventId?: string — Optional Cvent event ID filter
  • email?: string — Exact-match filter on attendee (contact) email — use for per-user scoping
  • status?: string — Filter by registration status (e.g. Accepted, Cancelled, Declined, Waitlisted, No Response)
  • modifiedSince?: string — Filter attendees added or updated on/after this ISO date
  • limit?: number — Max results (default 50, cap 200)
  • cursor?: string — Cvent paging token from a prior page's paging.nextToken
  • maxPages?: number — REST pages to walk in this one tool call (default 80, cap 80). Loop using cursor=paging.nextToken until hasMore is false.
JSON passthrough from provider APIparam: `email` (caller-asserted)
cvent_register_attendee
Register an attendee for an event. Resolves the Cvent contact by email (creating one if none exists), then adds the attendee to the event.
Writes on behalf of anyone — no linkage to a Nunc user. Resolves/creates the Cvent contact by email first.
  • eventId: string — Cvent event ID
  • firstName: string — Attendee first name
  • lastName: string — Attendee last name
  • email: string — Attendee email
  • company?: string — Attendee company
  • sendEmail?: boolean — Send the Cvent registration email to the attendee
JSON passthrough from provider APIparam: `email` (caller-asserted)
cvent_get_attendee_registrations
List a single attendee's registrations across all events, looked up by exact email — the per-user scoping read for app surfaces
THE per-user read for app surfaces — one attendee's registrations across all events by exact email.
  • email: string — Attendee email (exact match)
  • status?: string — Filter by registration status (e.g. Accepted, Pending, Cancelled)
  • limit?: number — Max results (default 50)
JSON passthrough from provider APIparam: `email` (caller-asserted)
cvent_cancel_registration
Cancel an attendee registration (sets the attendee's status to Cancelled)
  • eventId: string — Cvent event ID
  • attendeeId: string — Attendee registration ID
JSON passthrough from provider APIparam: `attendeeId` (caller-asserted)
cvent_checkin_attendee
Check in an attendee for an event. Updates attendance / check-in status and timestamp in Cvent.
  • eventId: string — Cvent event ID
  • attendeeId: string — Attendee registration ID
  • checkedInAt?: string — Check-in ISO datetime (defaults to current time)
JSON passthrough from provider APIparam: `attendeeId` (caller-asserted)
cvent_list_contacts
List contacts from Cvent's global contacts master directory
  • email?: string — Filter by contact email
  • modifiedSince?: string — ISO datetime for records added/updated after this date
  • limit?: number — Max results (default 50)
  • cursor?: string — Cvent paging token from a prior page's paging.nextToken
JSON passthrough from provider APIparam: `email` (caller-asserted)
cvent_get_contact
Get full contact details by Cvent contact ID
  • contactId: string — Cvent contact ID
JSON passthrough from provider APIparam: `contactId` (caller-asserted)
cvent_list_sessions
List sessions within an event
  • eventId: string — Cvent event ID
  • limit?: number — Max results (default 50)
JSON passthrough from provider API—
cvent_create_session
Create a session within an event
  • eventId: string — Cvent event ID
  • title: string — Session title
  • startTime: string — Start time (ISO 8601)
  • endTime: string — End time (ISO 8601)
  • description?: string — Session description
  • capacity?: number — Session capacity
JSON passthrough from provider API—
cvent_list_speakers
List speakers for an event in Cvent
  • eventId: string — Cvent event ID
  • limit?: number — Max results (default 50)
JSON passthrough from provider API—
cvent_get_speaker
Get speaker details by speaker ID
  • speakerId: string — Cvent speaker ID
JSON passthrough from provider API—
cvent_get_event_metrics
Get event engagement metrics from Attendee Insights — attendee count, activity count, average engagement score. Note: Requires that Attendee Insights engagement scoring is configured for the event in Cvent. If not configured, returns 'No Attendee Insights engagement score is configured'.
  • eventId: string — Cvent event ID
JSON passthrough from provider API—
cvent_list_engagement_scores
List per-attendee engagement scores (points + rank) for an event from Attendee Insights. Note: Requires that Attendee Insights engagement scoring is configured for the event in Cvent.
  • eventId: string — Cvent event ID
  • limit?: number — Max results (default 50)
JSON passthrough from provider API—
cvent_list_event_emails
List the emails configured for an event (use an email's ID with cvent_send_event_email)
  • eventId: string — Cvent event ID
  • limit?: number — Max results (default 50)
JSON passthrough from provider API—
cvent_send_event_email
Send one of the event's configured emails to specific attendees (asynchronous — returns per-attendee responses; check delivery with cvent_get_email_status)
Sends an event's configured email to attendee IDs — Guardian egress applies.
  • eventId: string — Cvent event ID
  • emailId: string — Event email ID (from cvent_list_event_emails)
  • attendeeIds: array<string> — Attendee IDs to send the email to
  • resendToPreviousRecipients?: boolean — Also resend to attendees who already received this email (default false)
JSON passthrough from provider APIparam: `emailId`, `attendeeIds` (caller-asserted)
cvent_get_email_status
Get delivery status for a previously sent event email request
  • emailRequestId: string — Email request ID returned when the email was sent
JSON passthrough from provider API—
cvent_list_seatings
List seating layouts for an event (read-only)
  • eventId: string — Cvent event ID
JSON passthrough from provider API—
cvent_get_seating
Get one seating layout by ID (read-only)
  • eventId: string — Cvent event ID
  • seatingId: string — Seating layout ID
JSON passthrough from provider API—
cvent_list_tables
List tables within a seating layout (read-only)
  • eventId: string — Cvent event ID
  • seatingId: string — Seating layout ID
  • limit?: number — Max results per page (cap 200)
  • cursor?: string — Cvent paging token from a prior page's paging.nextToken
  • maxPages?: number — REST pages to walk in this one tool call (default 10, cap 80).
JSON passthrough from provider API—
cvent_list_seats
List seats at a table within a seating layout (read-only)
  • eventId: string — Cvent event ID
  • seatingId: string — Seating layout ID
  • tableId: string — Unique alphanumeric identifier of the seating table within the event diagram
  • limit?: number — Max results per page (cap 200)
  • cursor?: string — Cvent paging token from a prior page's paging.nextToken
  • maxPages?: number — REST pages to walk in this one tool call (default 10, cap 80).
JSON passthrough from provider API—
cvent_list_seating_assignments
List attendee seat assignments for an event — who is seated where. Optionally scope to one seating layout and/or one attendee (read-only)
  • eventId: string — Cvent event ID
  • seatingId?: string — Optional seating layout ID to scope to
  • attendeeId?: string — Optional attendee ID to filter by
  • limit?: number — Max results (default 100, cap 200)
  • cursor?: string — Cvent paging token from a prior page's paging.nextToken
  • maxPages?: number — REST pages to walk in this one tool call (default 10, cap 80).
JSON passthrough from provider APIparam: `attendeeId` (caller-asserted)
Cvent (Service) cvent-service · 30 tools · Organization credentials

Server-to-server Cvent integration for autonomous applications and service keys. Uses Cvent OAuth client credentials under a dedicated catalog scope (cvent_svc_*) so PBAC can grant application access while keeping interactive user permissions separate.

Transport: cloud · Required credentials: CVENT_CLIENT_ID, CVENT_CLIENT_SECRET, CVENT_BASE_URL, CVENT_AUTH_URL, CVENT_SCOPE

Server-to-server Cvent client-credentials OAuth registered under a dedicated catalog scope (`cvent_svc_*`) so PBAC can authorize application keys independently from interactive users. 30 tools across Cvent events, attendees, contacts, sessions, and registrations, including `cvent_svc_get_attendee_registrations`.

ToolParametersReturnsUser Scoping
cvent_svc_list_events
List Cvent events. Use this for "most recent event", "upcoming events", or to get an event id before cvent_get_event. Most recent / latest-starting: sort=start:DESC and a small limit (1–5). Upcoming: status=Upcoming (Cvent eventStatus; Active is accepted as an alias), startDateAfter=<today ISO 8601>, sort=start:ASC. Recently edited: modifiedSince=<ISO 8601>. Next page: cursor from paging.nextToken. Cvent only sorts on start — without sort the page is not recency-ordered. Returns raw JSON { data, paging }.
  • status?: string — Lifecycle eventStatus (Upcoming, Ongoing, Completed, Cancelled, Archived) or planning status (Pending, Accepted). Active is treated as Upcoming.
  • startDateAfter?: string — Filter events starting on/after this ISO datetime or YYYY-MM-DD date (e.g. 2026-07-01T00:00:00Z or 2026-07-01)
  • startDateBefore?: string — Filter events starting on/before this ISO datetime or YYYY-MM-DD date (e.g. 2026-07-31T23:59:59Z or 2026-07-31)
  • modifiedSince?: string — Filter events added or updated on/after this ISO date
  • sort?: string — Sort by event start time. start:DESC = most recent / latest-starting first (use with limit=1 for a single answer). start:ASC = soonest-starting first. Cvent only sorts on start.
  • cursor?: string — Cvent paging token from a prior page's paging.nextToken
  • limit?: number — Max results to return (default 25)
JSON passthrough from provider API—
cvent_svc_get_event
Get full Cvent event details by ID. If you only have a name or need the latest/upcoming event, call cvent_list_events first (sort=start:DESC or start:ASC) and use an id from data[].
  • eventId: string — Cvent event ID
JSON passthrough from provider API—
cvent_svc_rotate_token
Mint a fresh Cvent OAuth access token and cache it until expires_in (Cvent: 60 minutes). Call once at the start of a poll; data GETs reuse the cached bearer. Does not fetch events or attendees.
noneJSON passthrough from provider API—
cvent_svc_create_event
Create a new event in Cvent (asynchronous — returns an eventCreationId and status; poll cvent_get_event_creation_status until status is Created, then the event ID is returned there)
  • title: string — Event title
  • startDate: string — Start date/time (ISO 8601)
  • endDate: string — End date/time (ISO 8601)
  • timezone: string — Event timezone (e.g. America/New_York)
  • eventType: string — Event type (e.g. Meeting, Conference, Seminar, TrainingSession, Webinar, TradeShow, Dinner, Forum)
  • plannerEmail: string — Email of the Cvent planner user who owns the event
  • format?: string — Event format: In-person (default), Virtual, or Hybrid
  • language?: string — IETF language tag for the event (default en-US)
  • description?: string — Event description
  • capacity?: number — Max attendees
JSON passthrough from provider API—
cvent_svc_get_event_creation_status
Check the status of an asynchronous event creation (Processing | Created | Failed); returns the new event ID once Created
  • eventCreationId: string — Creation ID returned by cvent_create_event
JSON passthrough from provider API—
cvent_svc_update_event
Update an existing event (read-modify-write: fetches the event, merges your updates, PUTs the full resource — Cvent has no partial update)
  • eventId: string — Cvent event ID
  • updates: object — Fields to update, using Cvent's field names (title, start, end, description, capacity, timezone, status, ...)
JSON passthrough from provider API—
cvent_svc_list_attendees
List attendees globally or filtered by event ID, email, or status
  • eventId?: string — Optional Cvent event ID filter
  • email?: string — Exact-match filter on attendee (contact) email — use for per-user scoping
  • status?: string — Filter by registration status (e.g. Accepted, Cancelled, Declined, Waitlisted, No Response)
  • modifiedSince?: string — Filter attendees added or updated on/after this ISO date
  • limit?: number — Max results (default 50, cap 200)
  • cursor?: string — Cvent paging token from a prior page's paging.nextToken
  • maxPages?: number — REST pages to walk in this one tool call (default 80, cap 80). Loop using cursor=paging.nextToken until hasMore is false.
JSON passthrough from provider APIparam: `email` (caller-asserted)
cvent_svc_register_attendee
Register an attendee for an event. Resolves the Cvent contact by email (creating one if none exists), then adds the attendee to the event.
  • eventId: string — Cvent event ID
  • firstName: string — Attendee first name
  • lastName: string — Attendee last name
  • email: string — Attendee email
  • company?: string — Attendee company
  • sendEmail?: boolean — Send the Cvent registration email to the attendee
JSON passthrough from provider APIparam: `email` (caller-asserted)
cvent_svc_get_attendee_registrations
List a single attendee's registrations across all events, looked up by exact email — the per-user scoping read for app surfaces
  • email: string — Attendee email (exact match)
  • status?: string — Filter by registration status (e.g. Accepted, Pending, Cancelled)
  • limit?: number — Max results (default 50)
JSON passthrough from provider APIparam: `email` (caller-asserted)
cvent_svc_cancel_registration
Cancel an attendee registration (sets the attendee's status to Cancelled)
  • eventId: string — Cvent event ID
  • attendeeId: string — Attendee registration ID
JSON passthrough from provider APIparam: `attendeeId` (caller-asserted)
cvent_svc_checkin_attendee
Check in an attendee for an event. Updates attendance / check-in status and timestamp in Cvent.
  • eventId: string — Cvent event ID
  • attendeeId: string — Attendee registration ID
  • checkedInAt?: string — Check-in ISO datetime (defaults to current time)
JSON passthrough from provider APIparam: `attendeeId` (caller-asserted)
cvent_svc_list_contacts
List contacts from Cvent's global contacts master directory
  • email?: string — Filter by contact email
  • modifiedSince?: string — ISO datetime for records added/updated after this date
  • limit?: number — Max results (default 50)
  • cursor?: string — Cvent paging token from a prior page's paging.nextToken
JSON passthrough from provider APIparam: `email` (caller-asserted)
cvent_svc_get_contact
Get full contact details by Cvent contact ID
  • contactId: string — Cvent contact ID
JSON passthrough from provider APIparam: `contactId` (caller-asserted)
cvent_svc_list_sessions
List sessions within an event
  • eventId: string — Cvent event ID
  • limit?: number — Max results (default 50)
JSON passthrough from provider API—
cvent_svc_create_session
Create a session within an event
  • eventId: string — Cvent event ID
  • title: string — Session title
  • startTime: string — Start time (ISO 8601)
  • endTime: string — End time (ISO 8601)
  • description?: string — Session description
  • capacity?: number — Session capacity
JSON passthrough from provider API—
cvent_svc_list_speakers
List speakers for an event in Cvent
  • eventId: string — Cvent event ID
  • limit?: number — Max results (default 50)
JSON passthrough from provider API—
cvent_svc_get_speaker
Get speaker details by speaker ID
  • speakerId: string — Cvent speaker ID
JSON passthrough from provider API—
cvent_svc_get_event_metrics
Get event engagement metrics from Attendee Insights — attendee count, activity count, average engagement score. Note: Requires that Attendee Insights engagement scoring is configured for the event in Cvent. If not configured, returns 'No Attendee Insights engagement score is configured'.
  • eventId: string — Cvent event ID
JSON passthrough from provider API—
cvent_svc_list_engagement_scores
List per-attendee engagement scores (points + rank) for an event from Attendee Insights. Note: Requires that Attendee Insights engagement scoring is configured for the event in Cvent.
  • eventId: string — Cvent event ID
  • limit?: number — Max results (default 50)
JSON passthrough from provider API—
cvent_svc_list_event_emails
List the emails configured for an event (use an email's ID with cvent_send_event_email)
  • eventId: string — Cvent event ID
  • limit?: number — Max results (default 50)
JSON passthrough from provider API—
cvent_svc_send_event_email
Send one of the event's configured emails to specific attendees (asynchronous — returns per-attendee responses; check delivery with cvent_get_email_status)
  • eventId: string — Cvent event ID
  • emailId: string — Event email ID (from cvent_list_event_emails)
  • attendeeIds: array<string> — Attendee IDs to send the email to
  • resendToPreviousRecipients?: boolean — Also resend to attendees who already received this email (default false)
JSON passthrough from provider APIparam: `emailId`, `attendeeIds` (caller-asserted)
cvent_svc_get_email_status
Get delivery status for a previously sent event email request
  • emailRequestId: string — Email request ID returned when the email was sent
JSON passthrough from provider API—
cvent_svc_list_seatings
List seating layouts for an event (read-only)
  • eventId: string — Cvent event ID
JSON passthrough from provider API—
cvent_svc_get_seating
Get one seating layout by ID (read-only)
  • eventId: string — Cvent event ID
  • seatingId: string — Seating layout ID
JSON passthrough from provider API—
cvent_svc_list_tables
List tables within a seating layout (read-only)
  • eventId: string — Cvent event ID
  • seatingId: string — Seating layout ID
  • limit?: number — Max results per page (cap 200)
  • cursor?: string — Cvent paging token from a prior page's paging.nextToken
  • maxPages?: number — REST pages to walk in this one tool call (default 10, cap 80).
JSON passthrough from provider API—
cvent_svc_list_seats
List seats at a table within a seating layout (read-only)
  • eventId: string — Cvent event ID
  • seatingId: string — Seating layout ID
  • tableId: string — Unique alphanumeric identifier of the seating table within the event diagram
  • limit?: number — Max results per page (cap 200)
  • cursor?: string — Cvent paging token from a prior page's paging.nextToken
  • maxPages?: number — REST pages to walk in this one tool call (default 10, cap 80).
JSON passthrough from provider API—
cvent_svc_list_seating_assignments
List attendee seat assignments for an event — who is seated where. Optionally scope to one seating layout and/or one attendee (read-only)
  • eventId: string — Cvent event ID
  • seatingId?: string — Optional seating layout ID to scope to
  • attendeeId?: string — Optional attendee ID to filter by
  • limit?: number — Max results (default 100, cap 200)
  • cursor?: string — Cvent paging token from a prior page's paging.nextToken
  • maxPages?: number — REST pages to walk in this one tool call (default 10, cap 80).
JSON passthrough from provider APIparam: `attendeeId` (caller-asserted)
cvent_svc_query_events
Incremental, read-only bulk EVENT pull for the events sync feed. Cursor-paginated via Cvent's paging token; loop until nextCursor is null. modifiedSince returns records added or updated after that time. Service/cron lane — no per-user identity argument.
  • eventId?: string — Optional Cvent event ID filter — scope bulk sync to one event.
  • status?: string — Optional registration status filter (e.g. Accepted, No Response, Cancelled).
  • modifiedSince?: string — ISO-8601 datetime, e.g. 2026-07-23T00:00:00Z. Returns records added or updated after this value (Cvent's `after` parameter). Omit for a full backfill. Ignored when a cursor is supplied.
  • limit?: number — Page size (default 200, capped at Cvent's max of 200).
  • cursor?: string — Opaque cursor from a prior page's nextCursor (Cvent's paging token); loop until nextCursor is null.
  • maxPages?: number — REST pages to walk in a single sync call (default 1, cap 80). One Cvent token exchange covers the entire page walk.
  • deadlineMs?: number — Max wall-clock milliseconds to spend walking pages before cleanly yielding accumulated records and nextCursor (default 22000, cap 28000).
JSON passthrough from provider API—
cvent_svc_query_contacts
Incremental, read-only bulk CONTACT pull (global Cvent contact/invitee master) for correlating Cvent people to members. Cursor-paginated; loop until nextCursor is null. modifiedSince returns records added or updated after that time. Service/cron lane — no per-user identity argument.
  • eventId?: string — Optional Cvent event ID filter — scope bulk sync to one event.
  • status?: string — Optional registration status filter (e.g. Accepted, No Response, Cancelled).
  • modifiedSince?: string — ISO-8601 datetime, e.g. 2026-07-23T00:00:00Z. Returns records added or updated after this value (Cvent's `after` parameter). Omit for a full backfill. Ignored when a cursor is supplied.
  • limit?: number — Page size (default 200, capped at Cvent's max of 200).
  • cursor?: string — Opaque cursor from a prior page's nextCursor (Cvent's paging token); loop until nextCursor is null.
  • maxPages?: number — REST pages to walk in a single sync call (default 1, cap 80). One Cvent token exchange covers the entire page walk.
  • deadlineMs?: number — Max wall-clock milliseconds to spend walking pages before cleanly yielding accumulated records and nextCursor (default 22000, cap 28000).
JSON passthrough from provider API—
cvent_svc_query_registrations
Incremental, read-only bulk REGISTRATION pull (cross-event attendance) — the core member-participation feed. Cursor-paginated; loop until nextCursor is null. modifiedSince returns records added or updated after that time. Service/cron lane — no per-user identity argument.
  • eventId?: string — Optional Cvent event ID filter — scope bulk sync to one event.
  • status?: string — Optional registration status filter (e.g. Accepted, No Response, Cancelled).
  • modifiedSince?: string — ISO-8601 datetime, e.g. 2026-07-23T00:00:00Z. Returns records added or updated after this value (Cvent's `after` parameter). Omit for a full backfill. Ignored when a cursor is supplied.
  • limit?: number — Page size (default 200, capped at Cvent's max of 200).
  • cursor?: string — Opaque cursor from a prior page's nextCursor (Cvent's paging token); loop until nextCursor is null.
  • maxPages?: number — REST pages to walk in a single sync call (default 1, cap 80). One Cvent token exchange covers the entire page walk.
  • deadlineMs?: number — Max wall-clock milliseconds to spend walking pages before cleanly yielding accumulated records and nextCursor (default 22000, cap 28000).
JSON passthrough from provider API—

finance (8)

Bill.com billcom · 11 tools · Organization credentials

Bill.com integration for accounts payable automation, invoice processing, payment approvals, and vendor management via REST API.

Transport: cloud + local (stdio) · Required credentials: BILLCOM_DEV_KEY, BILLCOM_USERNAME, BILLCOM_PASSWORD, BILLCOM_ORG_ID

ToolParametersReturnsUser Scoping
billcom_list_bills
List bills (accounts payable) with optional filters for status, vendor, date range, and amount.
  • status?: string — Filter by status (e.g., "1" = Open, "2" = Partial, "4" = Paid)
  • vendorId?: string — Filter by vendor ID
  • startDate?: string — Filter bills with due date on or after this date (YYYY-MM-DD)
  • endDate?: string — Filter bills with due date on or before this date (YYYY-MM-DD)
  • minAmount?: number — Minimum bill amount
  • maxAmount?: number — Maximum bill amount
  • start?: number — Pagination start index (default 0)
  • max?: number — Max results to return (default 50)
JSON passthrough from provider API—
billcom_list_pending_approvals
List bills currently waiting for approval by the calling user (or a specified approver email) by resolving their email to a Bill.com User ID and querying BillApprover workflow records.
  • approverEmail?: string — Optional approver email address to inspect. Defaults to the authenticated Nunc user's email.
  • includeUpcoming?: boolean — When true, also includes bills where the user is an upcoming approver (status = '4') in a multi-step approval chain.
  • max?: number — Max pending approvals to return (default 50)
JSON passthrough from provider API—
billcom_get_bill
Get a single bill by ID, including line items, full details, and the BillApprover workflow chain.
  • id: string — Unique alphanumeric identifier of the Bill.com bill record (e.g. '00n01000000000000001')
JSON passthrough from provider API—
billcom_create_bill
Create a new bill (accounts payable).
  • vendorId: string — Unique alphanumeric identifier of the vendor entity in Bill.com
  • invoiceNumber?: string — Vendor invoice number
  • invoiceDate?: string — Invoice date (YYYY-MM-DD)
  • dueDate: string — Payment due date (YYYY-MM-DD)
  • billLineItems: array<object> — Array of line item expense breakdowns detailing amount, chart of accounts, and description
  • description?: string — Bill description / memo
JSON passthrough from provider API—
billcom_approve_bill
Approve a bill for payment after verifying the calling Nunc user's Bill.com userId has active approval access for that specific bill.
  • id: string — Bill ID to approve
  • comment?: string — Optional approval note or comment recorded on the bill in Bill.com
JSON passthrough from provider API—
billcom_pay_bill
Schedule payment for an approved bill.
  • vendorId: string — Unique alphanumeric identifier of the vendor entity in Bill.com
  • billId: string — Bill ID to pay
  • amount: number — Payment amount
  • processDate: string — Date to process payment (YYYY-MM-DD)
  • bankAccountId: string — Bank account ID to pay from
JSON passthrough from provider API—
billcom_list_vendors
List vendors with optional search query.
  • name?: string — Search vendors by name
  • isActive?: string — Filter by active status ("1" = active, "2" = inactive)
  • start?: number — Pagination start index (default 0)
  • max?: number — Max results to return (default 50)
JSON passthrough from provider API—
billcom_create_vendor
Create a new vendor entity in Bill.com with contact and billing address details.
  • name: string — Vendor name
  • email?: string — Vendor email address
  • phone?: string — Vendor phone number
  • address1?: string — Street address line 1
  • address2?: string — Street address line 2
  • city?: string — City name for the vendor's physical mailing or billing address
  • state?: string — State (2-letter code)
  • zip?: string — Postal or ZIP code for the vendor's billing address
  • country?: string — Country (2-letter code, default US)
JSON passthrough from provider APIparam: `email` (caller-asserted)
billcom_list_invoices
List AR invoices (receivables / sent invoices).
  • customerId?: string — Filter by customer ID
  • status?: string — Filter by status
  • start?: number — Pagination start index (default 0)
  • max?: number — Max results to return (default 50)
JSON passthrough from provider API—
billcom_create_invoice
Create an AR invoice (sent invoice / receivable).
  • customerId: string — Unique alphanumeric identifier of the customer account in Bill.com
  • invoiceNumber?: string — Sequential billing invoice number assigned to this invoice (e.g. 'INV-2024-001')
  • invoiceDate?: string — Invoice date (YYYY-MM-DD)
  • dueDate: string — Payment due date (YYYY-MM-DD)
  • sentInvoiceLineItems: array<object> — Array of invoice line items detailing billable amount, quantity, and service description
  • description?: string — Invoice memo / description
JSON passthrough from provider API—
billcom_get_payment_status
Check the status of a bill payment.
  • id: string — BillPayment ID
JSON passthrough from provider API—
Expensify Hosted MCP expensify-mcp · runtime-discovered · Organization credentials

Federates Expensify's hosted MCP (expenses, reports, tags, categories, and SQL analytics) through Nunc /mcp. One authorized workspace Expensify account shared across Nunc users (OAuth 2.1 + PKCE with mcp:tools).

Transport: cloud · Required credentials: EXPENSIFY_MCP_URL, EXPENSIFY_CLIENT_ID, EXPENSIFY_CLIENT_SECRET, EXPENSIFY_AUTH_CODE, EXPENSIFY_REFRESH_TOKEN

Tools are discovered dynamically at runtime via tools/list once connected.

Sage Intacct intacct · 31 tools · Organization credentials

Sage Intacct ERP integration for general ledger, accounts payable/receivable, purchasing, and financial consolidation. Web Services API.

Transport: cloud + local (stdio) · Required credentials: INTACCT_COMPANY_ID, INTACCT_USER_ID, INTACCT_USER_PASSWORD, INTACCT_SENDER_ID, INTACCT_SENDER_PASSWORD, INTACCT_LOCATION_ID

ToolParametersReturnsUser Scoping
intacct_list_gl_accounts
List General Ledger accounts from Sage Intacct. Optionally filter by account type, status, or custom query.
  • filter?: string — Intacct query filter (e.g., "ACCOUNTTYPE = 'incomestatement'" or "STATUS = 'active'")
  • fields?: string — Comma-separated list of fields to return (default: all)
  • pageSize?: number — Number of records per page (default: 100, max: 1000)
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_get_gl_balance
Get the balance for a GL account across a specific reporting period or date range, including opening balance, period activity, and closing balance.
  • accountNo: string — GL account number (e.g., '1000', '1300')
  • reportingPeriod?: string — Reporting period name (e.g., "Quarter Ended September 2026", "Month Ended September 2026", "Current Month")
  • startDate?: string — Opening balance start date (e.g. '2026-07-01' or '07/01/2026'). Required if reportingPeriod is omitted.
  • endDate?: string — Closing balance end date (e.g. '2026-09-30' or '09/30/2026'). Required if reportingPeriod is omitted.
  • openingBalanceDate?: string — Alias for startDate (opening balance date)
  • closingBalanceDate?: string — Alias for endDate (closing balance date)
  • locationId?: string — Filter by location or entity ID (e.g. '100')
  • departmentId?: string — Filter by department ID
  • filter?: string — Additional query filter (optional)
JSON passthrough from provider API—
intacct_list_reporting_periods
List configured accounting reporting periods from Sage Intacct (e.g. Month Ended, Quarter Ended) with dates.
  • filter?: string — Optional query filter (e.g., "STATUS = 'Active'" or "budgeting = 'T'")
  • fields?: string — Comma-separated fields to return (default: all)
  • pageSize?: number — Number of records per page (default: 100, max: 1000)
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_create_journal_entry
Create an atomic journal entry (GL batch) in Sage Intacct. Enforces ACID validation requiring at least two balanced debit/credit line items.
  • journalSymbol: string — Journal symbol (e.g., "GJ" for General Journal)
  • datecreated: object — Date object with year, month, day (also accepts 'YYYY-MM-DD' string at runtime)
  • description?: string — Journal entry description/memo
  • state?: string — Posting state for the GL batch: 'Draft' (save as draft / unposted), 'Posted' (post immediately), or 'Submitted'. Defaults to 'Posted'.
  • draft?: boolean — Convenience flag: set to true to create the journal entry as a Draft (equivalent to state: 'Draft').
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100'). Automatically inferred when lines share a location / entity.
  • entityId?: string — Alias for locationId — Intacct entity/location ID to scope the API session (e.g. '100').
  • controlId?: string — Optional idempotency key / control ID to prevent duplicate creation on retry
  • lines: array<object> — Array of journal entry lines. Each must have accountno and either a signed amount (positive=debit, negative=credit) or separate debit/credit amounts. Total debits must equal total credits.
JSON passthrough from provider API—
intacct_list_gl_entries
Pull a General Ledger detail report or list posted General Ledger transaction lines (GLENTRY) from Sage Intacct for audit, reconciliation, and transaction drill-down. Supports filtering by GL account, date range, entity/location, department, journal symbol (e.g. 'GJ', 'CDJ', 'ARJ', 'APJ', 'CRJ'), batch/journal entry number (batchNo / entryNo, e.g. '1584' or 'GJ-1584'), document number (docNumber), record number (recordNo), customer, vendor, project, and class, with automatic account title, journal symbol, and debit/credit enrichment. Note: Sage Intacct numbers batches per journal symbol (so GJ #1584 and CDJ #1584 are distinct entries); pass journalSymbol (e.g. 'GJ') or a prefixed entryNo (e.g. 'GJ-1584') to target a specific journal.
  • accountNo?: string — Filter by GL account number (e.g. '1010', '4100', '2404')
  • journalSymbol?: string — Filter by journal symbol (e.g. 'GJ' for General Journal, 'CDJ' for Cash Disbursements Journal, 'ARJ' for AR Journal, 'APJ' for AP Journal, 'CRJ' for Cash Receipts Journal)
  • batchNo?: string — Filter by GL batch / journal entry number (BATCHNO, e.g. '1584' or 'GJ-1584'). Also falls back to DOCUMENT and RECORDNO if no BATCHNO matches.
  • entryNo?: string — Filter by journal entry / transaction number (e.g. '1584', '#1584', or 'GJ-1584'). Automatically searches GLBATCH (prioritizing General Journal 'GJ' when multiple journals share the same batch number), BATCHNO, DOCUMENT, and RECORDNO.
  • recordNo?: string — Filter by GLENTRY record number (RECORDNO)
  • docNumber?: string — Filter by document/reference number (DOCUMENT, e.g. '1584' or 'INV-00002802'). Also falls back to BATCHNO if no DOCUMENT matches.
  • customerId?: string — Filter by customer ID (e.g. 'C-000574')
  • vendorId?: string — Filter by vendor ID
  • projectId?: string — Filter by project ID (e.g. '3504')
  • departmentId?: string — Filter by department ID (e.g. '600')
  • classId?: string — Filter by class ID (e.g. '100')
  • state?: string — Filter by posting state (e.g. 'Posted', 'Draft')
  • startDate?: string — Filter entries on or after this date (YYYY-MM-DD)
  • endDate?: string — Filter entries on or before this date (YYYY-MM-DD)
  • filter?: string — Additional Intacct query filter expression
  • fields?: string — Comma-separated fields to return (default: all)
  • pageSize?: number — Maximum number of records to return (max 1000 per page). Omit to automatically paginate (via readMore) and return all matching GL entries.
  • locationId?: string — Optional entity/location ID to scope the session and filter GLENTRY rows (e.g. '100')
JSON passthrough from provider API—
intacct_get_journal_entry
Get a specific General Ledger transaction or journal entry (GLBATCH header and all GLENTRY debit/credit lines) from Sage Intacct by journal entry / batch number (entryNo / batchNo, e.g. '1584' or 'GJ-1584'), document number (docNumber), reference number, or RECORDNO. Because Sage Intacct numbers batches independently per journal ('GJ' General Journal vs 'CDJ' Cash Disbursements, 'ARJ' Accounts Receivable, 'APJ' Accounts Payable, 'CRJ' Cash Receipts), when journalSymbol is omitted and multiple journals share the same batch number, this tool prioritizes General Journal ('GJ') as the primary result and returns any other matching journals' batches in 'otherMatchingEntries'.
  • entryNo?: string — Journal entry or GL batch number, optionally prefixed with a journal symbol (e.g. '1584', '#1584', 'GJ-1584', 'GJ 1584', 'CDJ-1584')
  • batchNo?: string — Alias for entryNo — GL batch / journal entry number (BATCHNO, e.g. '1584' or 'GJ-1584')
  • recordNo?: string — Internal Intacct RECORDNO of the GLBATCH or GLENTRY
  • docNumber?: string — Document or reference number (DOCUMENT / REFERENCENO, e.g. '1584')
  • journalSymbol?: string — Optional journal symbol filter (e.g. 'GJ' for General Journal, 'CDJ' for Cash Disbursements Journal, 'ARJ' for AR Journal, 'APJ' for AP Journal, 'CRJ' for Cash Receipts Journal)
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_get_trial_balance
Get the comprehensive Trial Balance report from Sage Intacct for all accounts (cash, investments, AR, AP, deferred revenue, net assets, revenue, and expenses). Supports named reporting periods OR explicit opening/closing balance date ranges, entity/location filtering, zero-balance suppression, and pagination.
  • reportingPeriod?: string — Target accounting reporting period name (e.g. 'Quarter Ended September 2026', 'Month Ended September 2026', 'Current Month'). Required if startDate/endDate are omitted.
  • startDate?: string — Opening balance start date (e.g. '2026-07-01' or '07/01/2026'). Required if reportingPeriod is omitted.
  • endDate?: string — Closing balance end date (e.g. '2026-09-30' or '09/30/2026'). Required if reportingPeriod is omitted.
  • openingBalanceDate?: string — Alias for startDate (opening balance date).
  • closingBalanceDate?: string — Alias for endDate (closing balance date).
  • locationId?: string — Filter by location or entity ID (e.g. '100'). Required in multi-entity environments.
  • departmentId?: string — Filter by department ID
  • showZeroBalances?: boolean — Include accounts with zero balances and no activity. Default: false.
  • showDeptDetail?: boolean — Expand department dimension breakdown in the report. Default: false.
  • showLocDetail?: boolean — Expand location dimension breakdown in the report. Default: false.
  • includeSubs?: boolean — Include child locations / sub-entities when locationId is specified. Default: false (quarantines data to the specified entity).
  • reportingBook?: string — Optional reporting book ID (e.g. 'ACCRUAL', 'CASH'). Uses the company default book if omitted.
  • statistical?: string — Include statistical accounts ('include', 'exclude', or 'only').
  • pageSize?: number — Number of account rows to return per page. If omitted, returns all accounts.
  • pageNumber?: number — Page number to return when pageSize is provided (default: 1).
JSON passthrough from provider API—
intacct_list_bills
List Accounts Payable bills from Sage Intacct.
  • filter?: string — Query filter (e.g., "VENDORID = 'V-001'" or "STATE = 'Submitted'")
  • fields?: string — Comma-separated fields to return
  • pageSize?: number — Records per page (default: 100)
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_get_bill
Get a single Accounts Payable bill header (APBILL) along with all of its line items (APBILLITEM) by RECORDNO or bill number (RECORDID).
  • recordNo?: string — Internal Intacct RECORDNO of the AP bill
  • billNo?: string — Bill number / RECORDID of the AP bill (used if recordNo is omitted)
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_create_bill
Create an Accounts Payable bill in Sage Intacct.
  • vendorId: string — Sage Intacct vendor account ID (e.g. 'V-10001')
  • datecreated: object — Creation date object (also accepts 'YYYY-MM-DD' string at runtime)
  • datedue: object — Payment due date structured date object specifying year, month, and day (also accepts 'YYYY-MM-DD' string at runtime)
  • description?: string — Bill description or memo
  • billNo?: string — Optional explicit bill number / RECORDID. Omit when Intacct auto-numbers AP bills.
  • docNumber?: string — Optional vendor invoice / reference document number
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100'). Automatically inferred when all lines share a locationid.
  • action?: string — Submission action: 'Submit' (default) or 'Draft'
  • controlId?: string — Optional idempotency key / control ID to prevent duplicate bill creation on retry
  • lines: array<object> — Bill line items
JSON passthrough from provider API—
intacct_approve_bill
Approve an Accounts Payable bill in Sage Intacct.
  • recordNo: string — The RECORDNO of the AP bill to approve
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_list_ap_payments
List Accounts Payable payments (APPYMT) from Sage Intacct with optional filters for vendor, document number, state, or date range.
  • vendorId?: string — Filter by vendor ID (e.g. 'V-00100')
  • docNumber?: string — Filter by check or EFT reference number (DOCNUMBER)
  • state?: string — Filter by payment state (e.g. 'Posted', 'Approved')
  • startDate?: string — Filter payments with payment date (WHENPAID) on or after YYYY-MM-DD
  • endDate?: string — Filter payments with payment date (WHENPAID) on or before YYYY-MM-DD
  • filter?: string — Additional Intacct query filter expression
  • fields?: string — Comma-separated fields to return
  • pageSize?: number — Records per page (max 1000)
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_create_ap_payment
Create an atomic Accounts Payable payment request in Sage Intacct to pay one or more vendor bills. Automatically resolves billNo (RECORDID) to RECORDNO in a pre-flight check before submitting.
  • vendorId: string — Sage Intacct vendor ID (e.g. 'V-00100')
  • bankAccountId?: string — Intacct checking account ID (FINANCIALENTITY) to pay from. Mutually exclusive with chargeCardId.
  • chargeCardId?: string — Intacct charge card ID to pay from. Mutually exclusive with bankAccountId.
  • paymentMethod: string — Payment method: 'Printed Check', 'EFT' (or 'ACH'), 'Cash', or 'Credit Card'
  • paymentDate: string — Payment date in 'YYYY-MM-DD' format (or { year, month, day } object)
  • paymentAmount?: number — Optional total payment amount. If provided, ACID validation verifies it equals the exact sum of bills[].amount.
  • docNumber?: string — Optional check number, EFT trace, or external payment reference
  • memo?: string — Optional payment memo
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
  • controlId?: string — Optional idempotency key / control ID to prevent duplicate payment creation
  • bills: array<object> — Array of AP bills to pay. Each item requires amount (> 0) and either recordNo or billNo.
JSON passthrough from provider API—
intacct_list_vendors
List vendors from Sage Intacct.
  • filter?: string — Query filter (e.g., "STATUS = 'active'")
  • fields?: string — Comma-separated fields to return
  • pageSize?: number — Maximum number of records to return (max 1000 per page). Omit to automatically paginate (via readMore) and return all matching GL entries.
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_create_vendor
Create a new vendor record in Sage Intacct.
  • vendorId?: string — Optional explicit VENDORID. Omit if Intacct auto-numbers vendors.
  • name: string — Vendor legal or display name
  • email?: string — Primary contact email address
  • phone?: string — Primary contact phone number
  • address1?: string — Mailing address street line 1
  • address2?: string — Mailing address street line 2
  • city?: string — Mailing address city
  • state?: string — Mailing address state or province
  • zip?: string — Mailing address postal / ZIP code
  • country?: string — Mailing address country
  • status?: string — Vendor status: 'active' (default) or 'inactive'
  • termName?: string — Payment terms name (e.g. 'Net 30')
  • taxId?: string — Vendor tax ID / EIN
  • form1099Type?: string — Optional 1099 form type (e.g. 'NEC', 'MISC')
  • comments?: string — Optional notes or comments
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
  • controlId?: string — Optional idempotency key / control ID
JSON passthrough from provider APIparam: `email` (caller-asserted)
intacct_update_vendor
Update an existing vendor record in Sage Intacct.
  • vendorId: string — Sage Intacct VENDORID to update
  • name?: string — Updated vendor name
  • email?: string — Updated contact email address
  • phone?: string — Updated contact phone number
  • address1?: string — Updated street line 1
  • address2?: string — Updated street line 2
  • city?: string — Updated city
  • state?: string — Updated state or province
  • zip?: string — Updated postal / ZIP code
  • country?: string — Updated country
  • status?: string — Updated status ('active' or 'inactive')
  • termName?: string — Updated payment terms name
  • taxId?: string — Updated tax ID / EIN
  • comments?: string — Updated notes or comments
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
  • controlId?: string — Optional idempotency key / control ID
JSON passthrough from provider APIparam: `email` (caller-asserted)
intacct_list_invoices
List Accounts Receivable invoices from Sage Intacct.
  • filter?: string — Query filter (e.g., "CUSTOMERID = 'C-001'")
  • fields?: string — Comma-separated fields to return
  • pageSize?: number — Maximum number of records to return (max 1000 per page). Omit to automatically paginate (via readMore) and return all matching GL entries.
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_get_invoice
Get a single Accounts Receivable invoice header (ARINVOICE) along with all of its line items (ARINVOICEITEM) by RECORDNO or invoice number (RECORDID, e.g. 'INV-00002802').
  • recordNo?: string — Internal Intacct RECORDNO of the AR invoice (e.g. '18320')
  • invoiceNo?: string — Invoice number / RECORDID (e.g. 'INV-00002802'). Used if recordNo is omitted.
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_create_invoice
Create an Accounts Receivable invoice in Sage Intacct.
  • customerId: string — Sage Intacct customer account ID (e.g. 'C-10001')
  • datecreated: object — Creation date object (also accepts 'YYYY-MM-DD' string at runtime)
  • datedue: object — Due date object (also accepts 'YYYY-MM-DD' string at runtime)
  • description?: string — Invoice description or memo
  • invoiceNo?: string — Optional explicit invoice number / RECORDID (e.g. 'INV-00002802'). Omit to use Intacct document sequencing or automatic sequential numbering.
  • docNumber?: string — Optional reference or PO document number (DOCNUMBER)
  • locationId?: string — Optional Intacct entity or location ID to scope the session and invoice header (e.g. '100' for a specific entity instead of Top Level). Automatically inferred when all lines share a locationid.
  • action?: string — Submission action: 'Submit' (default) or 'Draft'
  • controlId?: string — Optional idempotency key / control ID to prevent duplicate invoice creation on retry
  • lines: array<object> — Invoice line items
JSON passthrough from provider API—
intacct_list_ar_payments
List Accounts Receivable customer payments (ARPYMT) from Sage Intacct with optional filters for customer, document/reference number, state, or payment date range.
  • customerId?: string — Filter by customer ID (e.g. 'C-000574')
  • docNumber?: string — Filter by payment reference / check / EFT number (DOCNUMBER, e.g. 'PMT-29894')
  • state?: string — Filter by payment state (e.g. 'Posted')
  • startDate?: string — Filter payments with WHENPAID on or after this date (YYYY-MM-DD)
  • endDate?: string — Filter payments with WHENPAID on or before this date (YYYY-MM-DD)
  • filter?: string — Additional Intacct query filter expression
  • fields?: string — Comma-separated fields to return
  • pageSize?: number — Maximum number of records to return (max 1000 per page). Omit to automatically paginate (via readMore) and return all matching GL entries.
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_receive_payment
Record and apply an atomic Accounts Receivable payment (create_arpayment) against one or more customer invoices in Sage Intacct. Automatically resolves human-readable invoiceNo (e.g. 'INV-00002802') to internal RECORDNO in a pre-flight check and enforces ACID balance validation.
  • customerId: string — Sage Intacct customer ID (e.g. 'C-000574')
  • paymentMethod: string — Payment method: 'EFT' (or 'ACH'/'Wire'), 'Printed Check' (or 'Check'), 'Cash', 'Credit Card', or 'Online'
  • dateReceived: string — Date payment was received in 'YYYY-MM-DD' format (or { year, month, day } object)
  • bankAccountId?: string — Intacct checking account ID (FINANCIALENTITY) to deposit the payment into. Specify either bankAccountId or undepositedFundsAccount.
  • undepositedFundsAccount?: string — Undeposited Funds GL account number (e.g. '1050'). Specify either bankAccountId or undepositedFundsAccount.
  • paymentAmount?: number — Optional explicit total payment amount. If provided, ACID validation verifies it matches the exact sum of invoices[].amount.
  • refId?: string — Optional check number, EFT trace, or external payment reference (e.g. Salesforce 'PMT-29894')
  • overpayLocId?: string — Optional location ID for overpayment coding (defaults to locationId)
  • overpayDeptId?: string — Optional department ID for overpayment coding
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
  • controlId?: string — Optional idempotency key / control ID to prevent duplicate payment creation on retry
  • invoices: array<object> — Array of invoices to apply this payment against. Each entry requires amount (> 0) and either recordNo (e.g. '18320') or invoiceNo (e.g. 'INV-00002802').
JSON passthrough from provider API—
intacct_create_ar_adjustment
Create an atomic Accounts Receivable credit memo, debit memo, or adjustment (create_aradjustment) for a customer in Sage Intacct.
  • customerId: string — Sage Intacct customer ID (e.g. 'C-000574')
  • datecreated: string — Adjustment creation date in 'YYYY-MM-DD' format (or { year, month, day } object)
  • adjustmentNo?: string — Optional adjustment number / ID (omit if Intacct auto-numbers AR adjustments)
  • invoiceNo?: string — Optional related invoice number (e.g. 'INV-00002802') associated with this adjustment
  • description?: string — Adjustment description or memo
  • action?: string — Submission action: 'Submit' (default) or 'Draft'
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
  • controlId?: string — Optional idempotency key / control ID to prevent duplicate adjustment creation
  • lines: array<object> — Adjustment line items (negative amount for credit memo reducing AR, positive for debit memo)
JSON passthrough from provider API—
intacct_list_customers
List customers from Sage Intacct.
  • filter?: string — Query filter (e.g., "STATUS = 'active'")
  • fields?: string — Comma-separated fields to return
  • pageSize?: number — Maximum number of records to return (max 1000 per page). Omit to automatically paginate (via readMore) and return all matching GL entries.
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_create_customer
Create a new customer record in Sage Intacct.
  • customerId?: string — Optional explicit CUSTOMERID (e.g. 'C-001200'). Omit if Intacct auto-numbers customers.
  • name: string — Customer account name
  • email?: string — Primary billing contact email address
  • phone?: string — Primary contact phone number
  • address1?: string — Billing address street line 1
  • address2?: string — Billing address street line 2
  • city?: string — Billing address city
  • state?: string — Billing address state or province
  • zip?: string — Billing address postal / ZIP code
  • country?: string — Billing address country
  • status?: string — Customer status: 'active' (default) or 'inactive'
  • termName?: string — Payment terms name (e.g. 'Net 30')
  • creditLimit?: number — Optional credit limit
  • taxId?: string — Optional customer tax ID
  • comments?: string — Optional notes or comments
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
  • controlId?: string — Optional idempotency key / control ID
JSON passthrough from provider APIparam: `email` (caller-asserted)
intacct_update_customer
Update an existing customer record in Sage Intacct.
  • customerId: string — Sage Intacct CUSTOMERID to update (e.g. 'C-000574')
  • name?: string — Updated customer name
  • email?: string — Updated primary contact email address
  • phone?: string — Updated contact phone number
  • address1?: string — Updated street line 1
  • address2?: string — Updated street line 2
  • city?: string — Updated city
  • state?: string — Updated state or province
  • zip?: string — Updated postal / ZIP code
  • country?: string — Updated country
  • status?: string — Updated status ('active' or 'inactive')
  • termName?: string — Updated payment terms name
  • creditLimit?: number — Updated credit limit
  • comments?: string — Updated comments
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
  • controlId?: string — Optional idempotency key / control ID
JSON passthrough from provider APIparam: `email` (caller-asserted)
intacct_list_checking_accounts
List Cash Management checking/bank accounts (CHECKINGACCOUNT) from Sage Intacct to discover valid bankAccountId (FINANCIALENTITY) values for AR and AP payments.
  • status?: string — Filter by status (e.g. 'active' or 'inactive')
  • filter?: string — Optional Intacct query filter expression
  • fields?: string — Comma-separated fields to return
  • pageSize?: number — Maximum number of records to return (max 1000 per page). Omit to automatically paginate (via readMore) and return all matching GL entries.
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_list_projects
List projects from Sage Intacct.
  • filter?: string — Query filter (e.g., "STATUS = 'active'")
  • fields?: string — Comma-separated fields to return
  • pageSize?: number — Maximum number of records to return (max 1000 per page). Omit to automatically paginate (via readMore) and return all matching GL entries.
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_list_departments
List departments from Sage Intacct.
  • filter?: string — Intacct SQL-like expression filter to constrain retrieved rows (e.g. "STATUS = 'active'")
  • fields?: string — Comma-separated fields to return
  • pageSize?: number — Maximum number of records to return (max 1000 per page). Omit to automatically paginate (via readMore) and return all matching GL entries.
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_list_locations
List locations/entities from Sage Intacct.
  • filter?: string — Intacct SQL-like expression filter to constrain retrieved rows (e.g. "STATUS = 'active'")
  • fields?: string — Comma-separated fields to return
  • pageSize?: number — Maximum number of records to return (max 1000 per page). Omit to automatically paginate (via readMore) and return all matching GL entries.
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_list_items
List Order Entry / AR items (ITEM) from Sage Intacct (e.g. '41010' for 'Calendar Year') so line items can reference valid ITEMID values.
  • status?: string — Filter by status (e.g. 'active' or 'inactive')
  • name?: string — Filter by item name (e.g. 'Calendar Year')
  • itemType?: string — Filter by item type
  • filter?: string — Optional Intacct query filter expression
  • fields?: string — Comma-separated fields to return
  • pageSize?: number — Maximum number of records to return (max 1000 per page). Omit to automatically paginate (via readMore) and return all matching GL entries.
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
intacct_list_classes
List classes (CLASS dimension, e.g. '100' for 'Unrestricted') from Sage Intacct.
  • status?: string — Filter by status (e.g. 'active' or 'inactive')
  • name?: string — Filter by class name (e.g. 'Unrestricted')
  • filter?: string — Optional Intacct query filter expression
  • fields?: string — Comma-separated fields to return
  • pageSize?: number — Maximum number of records to return (max 1000 per page). Omit to automatically paginate (via readMore) and return all matching GL entries.
  • locationId?: string — Optional entity/location ID to scope the Intacct API session (e.g. '100')
JSON passthrough from provider API—
QuickBooks quickbooks · 4 tools · Organization credentials

QuickBooks Online integration for invoices, expenses, customers, vendors, and financial reporting. OAuth2 authentication with Intuit API.

Transport: cloud + local (stdio) · Required credentials: QUICKBOOKS_ACCESS_TOKEN

ToolParametersReturnsUser Scoping
qb_get_customer
Get customer details by ID
  • customerId: string — QuickBooks customer ID
  • realmId: string — QuickBooks company ID (realm)
JSON passthrough from provider API—
qb_create_invoice
Create a new customer sales invoice in QuickBooks Online with line items and due dates.
  • customerId: string — Customer ID for the invoice
  • lineItems: array<object> — Line items with itemId, quantity, and rate
  • dueDate?: string — Invoice due date in YYYY-MM-DD format
  • realmId: string — QuickBooks company ID (realm)
JSON passthrough from provider API—
qb_list_invoices
List recent customer sales invoices from QuickBooks Online with customer names and balance totals.
  • limit?: number — Max invoices to return
  • realmId: string — QuickBooks company ID (realm)
JSON passthrough from provider API—
qb_get_payment_status
Check payment status for an invoice
  • invoiceId: string — Invoice ID to check
  • realmId: string — QuickBooks company ID (realm)
JSON passthrough from provider API—
Salesforce Hosted MCP salesforce-mcp · runtime-discovered · Organization credentials

Federates Salesforce's official hosted MCP (Headless 360 / sobject-*) through Nunc /mcp for Claude, Chat, and Data Studio. External Client App OAuth with mcp_api. For background M2M sync on /service, use Salesforce (Service).

Transport: cloud · Required credentials: SALESFORCE_LOGIN_URL, SALESFORCE_MCP_URL, SALESFORCE_CLIENT_ID, SALESFORCE_CLIENT_SECRET, SALESFORCE_REFRESH_TOKEN

Tools are discovered dynamically at runtime via tools/list once connected.

Salesforce (Service) salesforce-service · 21 tools · Organization credentials

Salesforce CRM integration for accounts, contacts, opportunities, custom objects, and SOQL queries via service authentication.

Transport: cloud · Required credentials: SALESFORCE_SVC_CLIENT_ID, SALESFORCE_SVC_USERNAME, SALESFORCE_SVC_PRIVATE_KEY, SALESFORCE_SVC_LOGIN_URL, SALESFORCE_SVC_CONTACT_FIELDS

Server-to-server OAuth 2.0 JWT Bearer flow (Connected App + integration user), no interactive user session. 21 tools — SOQL/SOSL/describe, generic record CRUD, and owner-scoped contacts/accounts/leads/opportunities/cases/tasks. Designed for autonomous applications; scope via PBAC to dedicated application keys.

ToolParametersReturnsUser Scoping
sf_svc_query
Run a read-only SOQL query (single SELECT statement). The general-purpose read for objects and filters the dedicated tools don't cover.
  • soql: string — SOQL SELECT statement, e.g. SELECT Id, Name FROM Account WHERE OwnerId = '005...' LIMIT 25
JSON passthrough from provider API—
sf_svc_search
Full-text SOSL search across objects (FIND {term} IN ALL FIELDS ...)
  • term: string — Search term
  • objects?: array<string> — SObjects to return, e.g. ["Contact", "Account"] (default Contact, Account, Lead, Opportunity)
  • limit?: number — Max results per object (default 20)
JSON passthrough from provider API—
sf_svc_describe
Describe an sobject — field names, types, picklist values, requiredness
  • sobject: string — SObject API name (e.g. Contact)
JSON passthrough from provider API—
sf_svc_get_record
Fetch one record by Id, optionally limited to specific fields
  • sobject: string — SObject API name
  • id: string — 18-character (or 15-character) alphanumeric Salesforce record ID
  • fields?: array<string> — Fields to return (default: all)
JSON passthrough from provider API—
sf_svc_create_record
Create a record of any sobject type from a field map
  • sobject: string — SObject API name
  • fields: object — Field map, e.g. { LastName: "Doe", Email: "..." }
JSON passthrough from provider API—
sf_svc_update_record
Update a record by Id from a field map
  • sobject: string — SObject API name
  • id: string — 18-character (or 15-character) alphanumeric Salesforce record ID
  • fields: object — Fields to update
JSON passthrough from provider API—
sf_svc_delete_record
Permanently delete a Salesforce record by record Id. Requires appropriate deletion permissions.
  • sobject: string — SObject API name
  • id: string — 18-character (or 15-character) alphanumeric Salesforce record ID
JSON passthrough from provider API—
sf_svc_find_contact
Find a contact by exact email — the per-user identity lookup. Returns the same fixed member-directory field projection as sf_svc_query_contacts so JIT-on-login can write a full profile_json.
  • email: string — Email address (exact match)
JSON passthrough from provider APIparam: `email` (caller-asserted)
sf_svc_create_contact
Create a new Contact record in Salesforce with name, email, phone, and account association.
  • firstName?: string — Contact's given first name
  • lastName: string — Contact's surname or family name (required by Salesforce)
  • email?: string — Primary email address for the contact
  • phone?: string — Primary contact telephone number with country/area code
  • accountId?: string — Parent Account Id (18-character Salesforce identifier) to link this contact to
  • ownerId?: string — Filter to records owned by this Salesforce User Id — the per-user scoping filter (caller-asserted; derive it from a verified session)
JSON passthrough from provider APIparam: `email`, `ownerId` (caller-asserted)
sf_svc_update_contact
Update a contact by Id from a field map
  • contactId: string — Contact record Id
  • fields: object — Fields to update
JSON passthrough from provider APIparam: `contactId` (caller-asserted)
sf_svc_query_contacts
Incremental, read-only bulk contact pull with a fixed server-defined field projection across standard and configured Contact fields. Keyset-paginated on (SystemModstamp, Id); loop until nextCursor is null. Designed for automated directory sync without a per-user identity argument.
  • modifiedSince?: string — ISO-8601 datetime, e.g. 2026-07-23T00:00:00Z. Returns contacts with SystemModstamp > this value. Omit for a full backfill.
  • limit?: number — Page size (default 200, capped at 2000).
  • cursor?: string — Opaque keyset cursor from a prior page's nextCursor; loop until nextCursor is null.
  • activeOnly?: boolean — When true, restrict to Active__c = true (current members only).
  • includeLapsedSince?: string — ISO-8601 date or datetime. Keep Active__c = true plus contacts whose Member_Drop_Date__c or npo02__MembershipEndDate__c is on/after this day (whichever of those fields the org has). Omit to pull every contact. Ignored when activeOnly is true.
JSON passthrough from provider API—
sf_svc_list_accounts
List accounts, optionally filtered by owner or name search
  • ownerId?: string — Filter to records owned by this Salesforce User Id — the per-user scoping filter (caller-asserted; derive it from a verified session)
  • nameContains?: string — Substring match on Name
  • limit?: number — Max results (default 25)
JSON passthrough from provider APIparam: `ownerId` (caller-asserted)
sf_svc_create_lead
Create a new Lead record in Salesforce with contact details, company name, and lead source.
  • firstName: string — Lead's given first name
  • lastName: string — Lead's surname or family name (required by Salesforce)
  • email: string — Business email address for the prospective lead
  • company: string — Legal entity or business organization name for the prospective lead
  • phone?: string — Primary contact telephone number with country/area code
  • ownerId?: string — Filter to records owned by this Salesforce User Id — the per-user scoping filter (caller-asserted; derive it from a verified session)
JSON passthrough from provider APIparam: `email`, `ownerId` (caller-asserted)
sf_svc_list_leads
List leads, optionally filtered by owner and status
  • ownerId?: string — Filter to records owned by this Salesforce User Id — the per-user scoping filter (caller-asserted; derive it from a verified session)
  • status?: string — Lead Status (e.g. Open, Working)
  • limit?: number — Max results (default 25)
JSON passthrough from provider APIparam: `ownerId` (caller-asserted)
sf_svc_list_opportunities
List opportunities, optionally filtered by owner, stage, or account
  • ownerId?: string — Filter to records owned by this Salesforce User Id — the per-user scoping filter (caller-asserted; derive it from a verified session)
  • stage?: string — StageName filter
  • accountId?: string — Parent Account Id
  • limit?: number — Max results (default 25)
JSON passthrough from provider APIparam: `ownerId` (caller-asserted)
sf_svc_create_opportunity
Create a new Opportunity pipeline record in Salesforce with stage name, close date, and amount.
  • name: string — Opportunity name
  • stageName: string — StageName (e.g. Prospecting, Closed Won)
  • closeDate: string — YYYY-MM-DD
  • amount?: number — Estimated revenue amount
  • accountId?: string — Parent Account Id
  • ownerId?: string — Filter to records owned by this Salesforce User Id — the per-user scoping filter (caller-asserted; derive it from a verified session)
JSON passthrough from provider APIparam: `ownerId` (caller-asserted)
sf_svc_update_opportunity
Update an opportunity by Id from a field map
  • opportunityId: string — Opportunity record Id
  • fields: object — Fields to update
JSON passthrough from provider API—
sf_svc_list_cases
List support cases, optionally filtered by contact email, owner, or status
  • contactEmail?: string — Exact contact email — per-user scoping for app surfaces (caller-asserted)
  • ownerId?: string — Filter to records owned by this Salesforce User Id — the per-user scoping filter (caller-asserted; derive it from a verified session)
  • status?: string — Case Status (e.g. New, Closed)
  • limit?: number — Max results (default 25)
JSON passthrough from provider APIparam: `ownerId` (caller-asserted)
sf_svc_create_case
Open a support case; contactEmail resolves to the Contact automatically
  • subject: string — Case subject
  • description?: string — Case description
  • contactEmail?: string — Contact email to attach the case to (exact match)
  • priority?: string — Low | Medium | High
  • ownerId?: string — Filter to records owned by this Salesforce User Id — the per-user scoping filter (caller-asserted; derive it from a verified session)
JSON passthrough from provider APIparam: `ownerId` (caller-asserted)
sf_svc_create_task
Create a task (activity), optionally linked to a who/what record
  • subject: string — Task subject
  • dueDate?: string — YYYY-MM-DD (ActivityDate)
  • whoId?: string — Contact/Lead Id the task is about
  • whatId?: string — Account/Opportunity/Case Id
  • ownerId?: string — Filter to records owned by this Salesforce User Id — the per-user scoping filter (caller-asserted; derive it from a verified session)
  • status?: string — Not Started | In Progress | Completed
JSON passthrough from provider APIparam: `ownerId` (caller-asserted)
sf_svc_list_tasks
List tasks, optionally filtered by owner and status
  • ownerId?: string — Filter to records owned by this Salesforce User Id — the per-user scoping filter (caller-asserted; derive it from a verified session)
  • status?: string — Status filter (e.g. Not Started, In Progress, Completed)
  • limit?: number — Max results (default 25)
JSON passthrough from provider APIparam: `ownerId` (caller-asserted)
Stripe stripe · 10 tools · Organization credentials

Stripe payments integration for subscriptions, customer billing, payment intents, refunds, and financial metrics. Real-time balance and charge visibility.

Transport: cloud + local (stdio) · Required credentials: STRIPE_SECRET_KEY

ToolParametersReturnsUser Scoping
stripe_balance_get
Retrieve the current Stripe account available and pending balance funds.
noneJSON passthrough from provider API—
stripe_customers_list
List or search Stripe customers by email or limit.
  • email?: string — Filter customers matching an email address.
  • limit?: number — Number of customers to return (default 20, max 100).
JSON passthrough from provider APIparam: `email` (caller-asserted)
stripe_customer_get
Retrieve full details for a single Stripe customer by Customer ID (cus_...).
  • customer_id: string — Stripe customer ID (e.g. cus_12345).
JSON passthrough from provider API—
stripe_payment_intents_list
List recent payment intents / charges with status, amounts, and customer references.
  • customer_id?: string — Filter by customer ID.
  • limit?: number — Max results to return (default 20, max 100).
JSON passthrough from provider API—
stripe_invoices_list
List invoices filtered by customer, subscription, or status (paid, open, void).
  • customer_id?: string — Filter by customer ID.
  • status?: string — Filter by status ('paid', 'open', 'void', 'draft', 'uncollectible').
  • limit?: number — Max results to return (default 20, max 100).
JSON passthrough from provider API—
stripe_subscriptions_list
List member subscriptions and billing intervals.
  • customer_id?: string — Filter subscriptions by customer ID.
  • status?: string — Filter by status ('active', 'past_due', 'canceled', 'all').
  • limit?: number — Max results to return (default 20, max 100).
JSON passthrough from provider API—
stripe_invoices_create
Create a new draft or payable invoice for a Stripe customer. Action tool.
  • customer_id: string — Stripe customer ID (cus_...).
  • description?: string — Memo or description displayed on the invoice.
  • currency?: string — Three-letter ISO currency code (e.g. 'usd').
  • collection_method?: string — Either 'send_invoice' (email to customer) or 'charge_automatically'. Default 'send_invoice'.
  • days_until_due?: number — Number of days until the invoice is due (when collection_method is send_invoice).
  • auto_advance?: boolean — Controls whether Stripe will automatically advance and finalize the invoice.
  • idempotency_key?: string — Optional Stripe Idempotency-Key header value.
JSON passthrough from provider API—
stripe_invoice_items_create
Add a line item or charge to a customer's draft invoice. Action tool.
  • customer_id: string — Stripe customer ID (cus_...).
  • invoice_id?: string — Optional draft Invoice ID (in_...) to attach this item to directly.
  • amount: number — Amount in cents (positive integer, e.g. 2500 for $25.00).
  • currency?: string — Three-letter ISO currency code (default 'usd').
  • description?: string — Line item description.
  • idempotency_key?: string — Optional Stripe Idempotency-Key header value.
JSON passthrough from provider API—
stripe_invoices_send
Finalize and send an invoice to the customer for payment via email. Action tool.
  • invoice_id: string — Stripe Invoice ID to send (in_...).
  • idempotency_key?: string — Optional Stripe Idempotency-Key header value.
JSON passthrough from provider API—
stripe_refunds_create
Create a refund for a charge or payment intent. Action tool.
  • payment_intent?: string — Payment Intent ID to refund (pi_...).
  • charge?: string — Charge ID to refund (ch_...).
  • amount?: number — Amount to refund in cents (positive integer; omit for full refund).
  • reason?: string — Reason for refund ('duplicate', 'fraudulent', 'requested_by_customer').
  • idempotency_key?: string — Optional Stripe Idempotency-Key header value.
JSON passthrough from provider API—
Stripe (Service) stripe-service · 10 tools · Organization credentials

Server-to-server Stripe integration for backend applications and service keys. Dedicated restricted API key scope so PBAC can grant stripe_svc_* to autonomous applications (e.g. member billing, refunds, customer subscriptions) while keeping user-tier keys isolated.

Transport: cloud + local (stdio) · Required credentials: STRIPE_SVC_SECRET_KEY, STRIPE_SVC_WEBHOOK_SECRET

ToolParametersReturnsUser Scoping
stripe_svc_balance_get
Retrieve the current Stripe account available and pending balance funds.
noneJSON passthrough from provider API—
stripe_svc_customers_list
List or search Stripe customers by email or limit.
  • email?: string — Filter customers matching an email address.
  • limit?: number — Number of customers to return (default 20, max 100).
JSON passthrough from provider APIparam: `email` (caller-asserted)
stripe_svc_customer_get
Retrieve full details for a single Stripe customer by Customer ID (cus_...).
  • customer_id: string — Stripe customer ID (e.g. cus_12345).
JSON passthrough from provider API—
stripe_svc_payment_intents_list
List recent payment intents / charges with status, amounts, and customer references.
  • customer_id?: string — Filter by customer ID.
  • limit?: number — Max results to return (default 20, max 100).
JSON passthrough from provider API—
stripe_svc_invoices_list
List invoices filtered by customer, subscription, or status (paid, open, void).
  • customer_id?: string — Filter by customer ID.
  • status?: string — Filter by status ('paid', 'open', 'void', 'draft', 'uncollectible').
  • limit?: number — Max results to return (default 20, max 100).
JSON passthrough from provider API—
stripe_svc_subscriptions_list
List member subscriptions and billing intervals.
  • customer_id?: string — Filter subscriptions by customer ID.
  • status?: string — Filter by status ('active', 'past_due', 'canceled', 'all').
  • limit?: number — Max results to return (default 20, max 100).
JSON passthrough from provider API—
stripe_svc_invoices_create
Create a new draft or payable invoice for a Stripe customer. Action tool.
  • customer_id: string — Stripe customer ID (cus_...).
  • description?: string — Memo or description displayed on the invoice.
  • currency?: string — Three-letter ISO currency code (e.g. 'usd').
  • collection_method?: string — Either 'send_invoice' (email to customer) or 'charge_automatically'. Default 'send_invoice'.
  • days_until_due?: number — Number of days until the invoice is due (when collection_method is send_invoice).
  • auto_advance?: boolean — Controls whether Stripe will automatically advance and finalize the invoice.
  • idempotency_key?: string — Optional Stripe Idempotency-Key header value.
JSON passthrough from provider API—
stripe_svc_invoice_items_create
Add a line item or charge to a customer's draft invoice. Action tool.
  • customer_id: string — Stripe customer ID (cus_...).
  • invoice_id?: string — Optional draft Invoice ID (in_...) to attach this item to directly.
  • amount: number — Amount in cents (positive integer, e.g. 2500 for $25.00).
  • currency?: string — Three-letter ISO currency code (default 'usd').
  • description?: string — Line item description.
  • idempotency_key?: string — Optional Stripe Idempotency-Key header value.
JSON passthrough from provider API—
stripe_svc_invoices_send
Finalize and send an invoice to the customer for payment via email. Action tool.
  • invoice_id: string — Stripe Invoice ID to send (in_...).
  • idempotency_key?: string — Optional Stripe Idempotency-Key header value.
JSON passthrough from provider API—
stripe_svc_refunds_create
Create a refund for a charge or payment intent. Action tool.
  • payment_intent?: string — Payment Intent ID to refund (pi_...).
  • charge?: string — Charge ID to refund (ch_...).
  • amount?: number — Amount to refund in cents (positive integer; omit for full refund).
  • reason?: string — Reason for refund ('duplicate', 'fraudulent', 'requested_by_customer').
  • idempotency_key?: string — Optional Stripe Idempotency-Key header value.
JSON passthrough from provider API—

geospatial (8)

ESRI ArcGIS esri · 10 tools · Organization credentials

ESRI ArcGIS geospatial integration for map services, geocoding, spatial analysis, and feature layer management via ArcGIS REST API.

Transport: cloud + local (stdio) · Required credentials: ESRI_API_KEY

ToolParametersReturnsUser Scoping
esri_geocode
Forward geocode an address to latitude/longitude coordinates
  • address: string — Street address to geocode
  • city?: string — City or municipality name to resolve coordinates for
  • region?: string — State or region
  • countryCode?: string — ISO 3166-1 country code (e.g. "US")
JSON passthrough from provider API—
esri_reverse_geocode
Convert latitude/longitude coordinates to a street address
  • latitude: number — Latitude coordinate
  • longitude: number — Longitude coordinate
JSON passthrough from provider API—
esri_route
Calculate a route between two or more stops with distance and time
  • stops: array<object> — Array of stop points (minimum 2)
  • travelMode?: string — Transportation modality for network impedance calculation ('driving', 'walking', 'trucking')
JSON passthrough from provider API—
esri_service_area
Calculate the reachable area from a point within a given travel time
  • latitude: number — Center point latitude
  • longitude: number — Center point longitude
  • travelTime: number — Maximum travel time in minutes
  • travelMode?: string — Mode of travel for service area calculation
JSON passthrough from provider API—
esri_search_places
Search for places and points of interest near a location
  • searchText: string — Search query (e.g. "coffee shop", "hospital")
  • latitude: number — Search center latitude
  • longitude: number — Search center longitude
  • radius?: number — Search radius in meters
  • categories?: string — Comma-separated category filters
JSON passthrough from provider API—
esri_query_features
Query a feature service layer with SQL-like filters
  • serviceUrl: string — Full URL to the feature service layer
  • where?: string — SQL WHERE clause (e.g. "POP > 100000")
  • outFields?: string — Comma-separated field names to return
  • returnGeometry?: boolean — Whether to include geometry in results
JSON passthrough from provider API—
esri_spatial_analysis
Run spatial analysis operations like buffer, intersect, or dissolve
  • operation: string — Spatial operation to perform
  • inputGeojson: object — Input GeoJSON geometry or feature collection
  • params?: object — Operation-specific parameters (e.g. { distance: 1000, unit: "meters" } for buffer)
JSON passthrough from provider API—
esri_elevation
Get elevation data at one or more geographic points
  • points: array<object> — Array of points to query elevation for
JSON passthrough from provider API—
esri_demographics
Get demographic data (population, income, age, etc.) for an area
  • latitude: number — Center point latitude
  • longitude: number — Center point longitude
  • radius?: number — Study area radius in miles
  • dataCollections?: string — Comma-separated data collection IDs (e.g. "KeyGlobalFacts,KeyUSFacts")
JSON passthrough from provider API—
esri_create_map
Generate a web map JSON configuration for embedding
  • basemap?: string — Basemap style
  • layers?: array<object> — Array of operational layers to include
  • extent?: object — Map extent (bounding box)
  • title?: string — User-facing title for the web map displayed in the ArcGIS portal
JSON passthrough from provider API—
Geotab geotab · 6 tools · Organization credentials

Geotab fleet telematics (MyGeotab) — devices, real-time vehicle status and location, trips, rule exceptions, zones/geofences, and GPS logs.

Transport: cloud + local (stdio) · Required credentials: GEOTAB_USERNAME, GEOTAB_PASSWORD, GEOTAB_DATABASE, GEOTAB_SERVER

ToolParametersReturnsUser Scoping
geotab_list_devices
List vehicles/devices in the fleet (id, name, VIN, serial). Optionally filter by name.
  • name?: string — Filter by device name (exact or wildcard)
  • limit?: number — Max devices to return
JSON passthrough from provider API—
geotab_device_status
Get real-time status for vehicles — current location (lat/lon), speed, bearing, and whether the engine is on. Optionally for one device.
  • deviceId?: string — Device id to scope to (omit for whole fleet)
  • limit?: number — Max records
JSON passthrough from provider API—
geotab_get_trips
Get completed trips for a device within a date range (distance, duration, start/stop).
  • deviceId: string — Unique telematics device hardware ID installed in the vehicle (e.g. 'b1')
  • fromDate: string — ISO 8601 start datetime (UTC)
  • toDate: string — ISO 8601 end datetime (UTC)
  • limit?: number — Maximum number of trip records to return (min 1, max 1000, default 100)
JSON passthrough from provider API—
geotab_get_exceptions
Get rule-exception events (speeding, harsh braking, idling, etc.) within a date range, optionally for one device.
  • deviceId?: string — Device id (omit for whole fleet)
  • fromDate: string — ISO 8601 start datetime (UTC)
  • toDate: string — ISO 8601 end datetime (UTC)
  • limit?: number — Max events
JSON passthrough from provider API—
geotab_get_logs
Get raw GPS breadcrumb log records (lat/lon/speed over time) for a device within a date range.
  • deviceId: string — Unique telematics device hardware ID installed in the vehicle (e.g. 'b1')
  • fromDate: string — ISO 8601 start datetime (UTC)
  • toDate: string — ISO 8601 end datetime (UTC)
  • limit?: number — Max log records
JSON passthrough from provider API—
geotab_get_zones
List zones/geofences defined in the fleet. Optionally filter by name.
  • name?: string — Filter by zone name
  • limit?: number — Maximum number of geofence zone records to return (min 1, max 1000, default 100)
JSON passthrough from provider API—
HERE here · 6 tools · Organization credentials

HERE Location Services — geocoding, reverse geocoding, POI discovery, routing (multi-modal), and real-time traffic flow and incidents.

Transport: cloud + local (stdio) · Required credentials: HERE_API_KEY

ToolParametersReturnsUser Scoping
here_geocode
Forward geocode: convert an address or place name into coordinates.
  • query: string — Address or place to geocode
  • limit?: number — Max results
  • countryCode?: string — ISO 3166-1 alpha-3 country code(s) to restrict results, e.g. 'USA'
JSON passthrough from provider API—
here_reverse_geocode
Reverse geocode: convert coordinates into the nearest address.
  • latitude: number — WGS 84 geographic latitude coordinate in decimal degrees (range: -90.0 to 90.0)
  • longitude: number — WGS 84 geographic longitude coordinate in decimal degrees (range: -180.0 to 180.0)
  • limit?: number — Max results
JSON passthrough from provider API—
here_discover
Search for places/POIs near a location by free-text query.
  • query: string — Search text, e.g. 'restaurant'
  • latitude: number — Search center latitude
  • longitude: number — Search center longitude
  • limit?: number — Max results
JSON passthrough from provider API—
here_route
Calculate a route between origin and destination with summary (distance, duration) and polyline.
  • origin: object — Starting point coordinates (lat, lng)
  • destination: object — Ending destination coordinates (lat, lng)
  • transportMode?: string — Mode of transportation for routing
JSON passthrough from provider API—
here_traffic_flow
Get real-time traffic flow (speed, jam factor) within a radius of a point.
  • latitude: number — Center latitude
  • longitude: number — Center longitude
  • radius?: number — Radius in meters
JSON passthrough from provider API—
here_traffic_incidents
Get real-time traffic incidents (accidents, closures, road works) within a radius of a point.
  • latitude: number — Center latitude
  • longitude: number — Center longitude
  • radius?: number — Radius in meters
JSON passthrough from provider API—
HPMS hpms · 2 tools · Built-in (no external credentials)

MCP server for FHWA Highway Performance Monitoring System data.

Transport: cloud + local (stdio)

ToolParametersReturnsUser Scoping
hpms_get_aadt
Fetch AADT (Average Annual Daily Traffic) data for a US state from the FHWA HPMS database. Returns GeoJSON FeatureCollection of road LineStrings with AADT, ROUTE_ID, THROUGH_LANES, SPEED_LIMIT, URBAN_CODE properties.
  • state: string — Two-letter US state abbreviation (e.g. "CA", "NY", "TX")
  • route_id?: string — Optional route ID filter (e.g. "CA-001")
  • year?: number — Data year (default 2021). Available years vary by state.
  • min_aadt?: number — Minimum AADT threshold to filter results (default 0)
  • max_features?: number — Maximum number of features to return (default 500)
JSON passthrough from provider API—
hpms_get_routes
Fetch highway route geometries for a US state from the FHWA HPMS database. Returns GeoJSON FeatureCollection of route LineStrings with ROUTE_ID, ROUTE_NAME, F_SYSTEM, THROUGH_LANES properties.
  • state: string — Two-letter US state abbreviation (e.g. "CA", "NY", "TX")
  • route_type?: string — Filter by route type: interstate, us (US highways), or state (state routes)
  • max_features?: number — Maximum number of features to return (default 500)
JSON passthrough from provider API—
Mapbox mapbox · 6 tools · Organization credentials

Mapbox APIs — forward/reverse geocoding, directions with live traffic, isochrones (reachable areas), travel-time matrices, and tilequery feature lookups.

Transport: cloud + local (stdio) · Required credentials: MAPBOX_ACCESS_TOKEN

ToolParametersReturnsUser Scoping
mapbox_geocode
Forward geocode: convert an address or place name into coordinates (Geocoding v6).
  • query: string — Address or place to geocode
  • limit?: number — Max results (1-10)
  • country?: string — Comma-separated ISO 3166-1 country codes to limit results
  • types?: string — Comma-separated feature types, e.g. 'address,poi,place'
JSON passthrough from provider API—
mapbox_reverse_geocode
Reverse geocode: convert coordinates into place features (Geocoding v6).
  • latitude: number — WGS 84 geographic latitude coordinate in decimal degrees (range: -90.0 to 90.0)
  • longitude: number — WGS 84 geographic longitude coordinate in decimal degrees (range: -180.0 to 180.0)
  • types?: string — Comma-separated feature types to return
JSON passthrough from provider API—
mapbox_directions
Get a route between waypoints with distance, duration, and turn-by-turn geometry. Supports live-traffic driving.
  • waypoints: array<object> — Ordered waypoints (min 2)
  • profile?: string — Routing profile mode
JSON passthrough from provider API—
mapbox_isochrone
Compute isochrones — the area reachable from a point within given travel-time contours.
  • latitude: number — Center latitude
  • longitude: number — Center longitude
  • profile?: string — Travel mode profile for isochrone calculation
  • contoursMinutes: array<number> — Travel-time contours in minutes, e.g. [10, 20, 30]
JSON passthrough from provider API—
mapbox_matrix
Compute a travel-time/distance matrix between a set of coordinates (durations between every pair).
  • coordinates: array<object> — Points to compute the matrix over (max 25)
  • profile?: string — Routing profile mode for distance matrix
JSON passthrough from provider API—
mapbox_tilequery
Query vector-tile features from a tileset near a point (e.g. nearby roads, POIs, boundaries).
  • tilesetId?: string — Tileset id, e.g. 'mapbox.mapbox-streets-v8'
  • latitude: number — Query point latitude in WGS 84 decimal degrees (-90 to 90)
  • longitude: number — Query point longitude in WGS 84 decimal degrees (-180 to 180)
  • radius?: number — Search radius in meters
  • limit?: number — Max features (1-50)
JSON passthrough from provider API—
StreetLight Data streetlight · 6 tools · Organization credentials

StreetLight Data InSight — mobility analytics over zones: list and create analyses (origin-destination, zone activity, segment), poll status, fetch metrics, and manage zone sets.

Transport: cloud + local (stdio) · Required credentials: STREETLIGHT_API_KEY

ToolParametersReturnsUser Scoping
streetlight_list_analyses
List analyses in the account with their status (Pending, Processing, Available, Error).
  • status?: string — Filter by status, e.g. 'Available'
JSON passthrough from provider API—
streetlight_get_analysis
Get a single analysis by UUID — its configuration and current processing status.
  • uuid: string — Analysis UUID
JSON passthrough from provider API—
streetlight_create_analysis
Create a new analysis. Pass the full InSight analysis request body (analysis_name, analysis_type, travel_mode_type, date_ranges, zones/zone sets, etc.). Returns the new analysis UUID; poll streetlight_get_analysis until status is 'Available'.
  • analysis: object — InSight analysis request body. Minimum: analysis_name, analysis_type (e.g. 'OD_Analysis', 'Zone_Activity_Analysis', 'Segment_Analysis'), travel_mode_type, date_ranges, and zone definitions.
JSON passthrough from provider API—
streetlight_get_analysis_metrics
Fetch the computed metrics for an analysis once it is Available. Optionally select a specific metric type.
  • uuid: string — Analysis UUID
  • metricType?: string — Metric type to retrieve, e.g. 'od_all', 'za_all', 'sa_all' (defaults to all available)
JSON passthrough from provider API—
streetlight_list_zone_sets
List zone sets (collections of geographic zones) in the account.
noneJSON passthrough from provider API—
streetlight_create_zone_set
Create a zone set from a GeoJSON FeatureCollection of zones for use in analyses.
  • name: string — Zone set name
  • zones: object — GeoJSON FeatureCollection of zone geometries
  • isPassThrough?: boolean — Treat zones as pass-through (line) zones
JSON passthrough from provider API—
TIGER/Web tiger · 3 tools · Built-in (no external credentials)

MCP server for US Census Bureau TIGER/Web GIS data.

Transport: cloud + local (stdio)

ToolParametersReturnsUser Scoping
tiger_get_tracts
Fetch census tract boundaries from the US Census Bureau TIGER/Web service. Returns GeoJSON FeatureCollection of tract polygons with GEOID, NAME, ALAND, AWATER properties.
  • state_fips: string — Two-digit state FIPS code (e.g. "06" for California, "36" for New York)
  • county_fips?: string — Three-digit county FIPS code (optional, filters to a specific county)
  • year?: number — Census year (not currently used by TIGER/Web current layer, reserved for future use)
  • max_features?: number — Maximum number of features to return (default 500)
JSON passthrough from provider API—
tiger_get_counties
Fetch county boundaries for a US state from the Census Bureau TIGER/Web service. Returns GeoJSON FeatureCollection of county polygons with GEOID, NAME, ALAND, AWATER properties.
  • state_fips: string — Two-digit state FIPS code (e.g. "06" for California)
  • max_features?: number — Maximum number of features to return (default 500)
JSON passthrough from provider API—
tiger_get_roads
Fetch primary roads from the US Census Bureau TIGER/Web service. Returns GeoJSON FeatureCollection of road LineStrings with FULLNAME, MTFCC properties.
  • state_fips: string — Two-digit state FIPS code (e.g. "06" for California)
  • county_fips?: string — Three-digit county FIPS code (optional, filters to a specific county)
  • max_features?: number — Maximum number of features to return (default 500)
JSON passthrough from provider API—
TomTom tomtom · 6 tools · Organization credentials

TomTom Maps APIs — geocoding, reverse geocoding, POI search, routing with live traffic, and real-time traffic incidents and flow.

Transport: cloud + local (stdio) · Required credentials: TOMTOM_API_KEY

ToolParametersReturnsUser Scoping
tomtom_geocode
Forward geocode: convert an address or place name into latitude/longitude coordinates.
  • query: string — Address or place to geocode, e.g. '4700 Mueller Blvd, Austin TX'
  • limit?: number — Max results (1-100)
  • countrySet?: string — Comma-separated ISO country codes to bias results, e.g. 'US,CA'
JSON passthrough from provider API—
tomtom_reverse_geocode
Reverse geocode: convert latitude/longitude into a street address.
  • latitude: number — Geographic latitude in decimal degrees (WGS 84, range: -90.0 to 90.0)
  • longitude: number — Geographic longitude in decimal degrees (WGS 84, range: -180.0 to 180.0)
JSON passthrough from provider API—
tomtom_search_poi
Search for points of interest (businesses, landmarks, services) near a location.
  • query: string — POI search text, e.g. 'coffee', 'ev charging'
  • latitude: number — Search center latitude
  • longitude: number — Search center longitude
  • radius?: number — Search radius in meters
  • limit?: number — Max results (1-100)
  • categorySet?: string — Comma-separated TomTom category IDs to filter by
JSON passthrough from provider API—
tomtom_route
Calculate a route between an origin and destination with distance, travel time, and live traffic.
  • origin: object — Origin point
  • destination: object — Destination point
  • travelMode?: string — Transportation modality used to calculate speed profiles, legal restrictions, and routes
  • routeType?: string — Routing calculation mode
  • traffic?: boolean — Account for live traffic
JSON passthrough from provider API—
tomtom_traffic_incidents
Get real-time traffic incidents (jams, road works, closures, accidents) within a bounding box.
  • bbox: object — Bounding box (WGS84)
  • language?: string — Language for incident descriptions, e.g. 'en-US'
JSON passthrough from provider API—
tomtom_traffic_flow
Get current traffic flow (current vs. free-flow speed and travel time) at a point.
  • latitude: number — Point latitude
  • longitude: number — Point longitude
  • zoom?: number — Zoom level 0-22 (road granularity)
JSON passthrough from provider API—

hr (2)

BambooHR bamboohr · 10 tools · Organization credentials

BambooHR integration for employee records, time-off requests, onboarding workflows, and organizational reporting via BambooHR API.

Transport: cloud + local (stdio) · Required credentials: BAMBOOHR_API_KEY, BAMBOOHR_SUBDOMAIN

ToolParametersReturnsUser Scoping
bamboohr_list_employees
List all employees from the BambooHR directory
noneJSON passthrough from provider API—
bamboohr_get_employee
Get detailed information about a specific employee
  • id: string — Employee ID
  • fields?: string — Comma-separated field names (e.g. "firstName,lastName,department,jobTitle,workEmail,hireDate,status")
JSON passthrough from provider API—
bamboohr_search_employees
Search employees in the directory by name or other criteria
  • query: string — Search query to filter employees by name
JSON passthrough from provider API—
bamboohr_get_time_off
Get time-off requests for an employee
  • id: string — Employee ID
  • status?: string — Filter by status
  • start?: string — Start date (YYYY-MM-DD)
  • end?: string — End date (YYYY-MM-DD)
JSON passthrough from provider API—
bamboohr_request_time_off
Submit a time-off request for an employee
  • id: string — Employee ID
  • timeOffTypeId: string — Time-off type ID
  • start: string — Start date (YYYY-MM-DD)
  • end: string — End date (YYYY-MM-DD)
  • amount?: number — Number of days/hours requested per day
  • notes?: string — Request notes
JSON passthrough from provider API—
bamboohr_list_time_off_types
List available time-off types (vacation, sick, etc.)
noneJSON passthrough from provider API—
bamboohr_whos_out
Get a list of who is out on a given date range
  • start?: string — Start date (YYYY-MM-DD)
  • end?: string — End date (YYYY-MM-DD)
JSON passthrough from provider API—
bamboohr_get_org_chart
Get the organizational chart (employee hierarchy built from reportsTo)
noneJSON passthrough from provider API—
bamboohr_list_departments
List all departments derived from the employee directory
noneJSON passthrough from provider API—
bamboohr_get_employee_files
List files associated with an employee
  • id: string — Employee ID
JSON passthrough from provider API—
Greenhouse greenhouse · 12 tools · Organization credentials

Greenhouse ATS integration for job postings, candidate pipelines, interview scheduling, scorecards, and hiring analytics via Harvest API.

Transport: cloud + local (stdio) · Required credentials: GREENHOUSE_API_KEY, GREENHOUSE_ON_BEHALF_OF

ToolParametersReturnsUser Scoping
greenhouse_list_jobs
List jobs with optional filters
  • status?: string — Filter by job status
  • department_id?: number — Filter by department ID
  • per_page?: number — Results per page (max 500)
  • page?: number — 1-based page number to fetch for paginated listings
JSON passthrough from provider API—
greenhouse_get_job
Get detailed information about a specific job
  • id: number — Unique numeric job requisition ID in Greenhouse (e.g. 12345)
JSON passthrough from provider API—
greenhouse_create_job
Create a new job opening requisition in Greenhouse from a template job with department and title.
  • template_job_id: number — Numeric identifier of the Greenhouse template job to clone from
  • title: string — Official job title for the opening (e.g. 'Senior Staff Software Engineer')
  • department_id?: number — Numeric identifier of the hiring department
  • office_ids?: array<number> — Array of office location IDs where the role is based
  • number_of_openings?: number — Total headcount openings to recruit for this role
  • notes?: string — Internal recruitment notes and hiring requirements visible only to the hiring team
JSON passthrough from provider API—
greenhouse_list_candidates
List candidates with optional filters
  • job_id?: number — Filter by job ID
  • email?: string — Filter by email address
  • updated_after?: string — ISO 8601 timestamp to filter by last update
  • per_page?: number — Results per page (max 500)
  • page?: number — 1-based page number to fetch for paginated listings
JSON passthrough from provider APIparam: `email` (caller-asserted)
greenhouse_get_candidate
Get detailed information about a specific candidate
  • id: number — Numeric identifier of the candidate in Greenhouse
JSON passthrough from provider API—
greenhouse_add_candidate
Add a new candidate to Greenhouse
  • first_name: string — Candidate legal or preferred first name
  • last_name: string — Candidate's surname or family name
  • email?: string — Email address
  • phone?: string — Candidate primary telephone contact number
  • applications?: array<object> — Job applications to create
JSON passthrough from provider APIparam: `email` (caller-asserted)
greenhouse_list_applications
List applications with optional filters
  • job_id?: number — Filter by job ID
  • status?: string — Filter by status
  • candidate_id?: number — Filter by candidate ID
  • per_page?: number — Results per page (max 500)
  • page?: number — 1-based page number to fetch for paginated listings
JSON passthrough from provider API—
greenhouse_advance_application
Advance an application to the next interview stage
  • id: number — Numeric identifier of the application in Greenhouse
  • from_stage_id?: number — Current stage ID (optional)
JSON passthrough from provider API—
greenhouse_reject_application
Reject a candidate application in Greenhouse, recording the rejection reason.
  • id: number — Numeric identifier of the application in Greenhouse
  • rejection_reason_id?: number — Greenhouse rejection reason catalog identifier
  • notes?: string — Hiring decision rationale notes explaining the rejection
  • rejection_email?: object — Rejection email delivery configuration object
JSON passthrough from provider APIparam: `rejection_email` (caller-asserted)
greenhouse_list_scorecards
List scorecards for an application
  • application_id: number — Numeric identifier of the application whose interview scorecards to list
JSON passthrough from provider API—
greenhouse_list_departments
List all departments in Greenhouse
  • per_page?: number — Results per page (max 500)
  • page?: number — 1-based page number to fetch for paginated listings
JSON passthrough from provider API—
greenhouse_list_offices
List all offices in Greenhouse
  • per_page?: number — Results per page (max 500)
  • page?: number — 1-based page number to fetch for paginated listings
JSON passthrough from provider API—

local (1)

Nunc Local Tools nunc-local · 19 tools · Built-in (no external credentials)

Read, write, and manage local files, directories, and execute allowlisted commands. Provides sandboxed operations under the workspace root.

Transport: local (stdio)

ToolParametersReturnsUser Scoping
local_fs_read_file
Read file contents from the workspace. Returns text (utf-8) or base64 string for binary files. Maximum file size 10MB.
  • path: string — Target file or directory path relative to the workspace root directory
  • encoding?: string — Data encoding: 'utf-8' returns readable text, 'base64' encodes binary data
JSON passthrough from provider API—
local_fs_write_file
Write textual or binary content to a file in the workspace. Automatically creates parent directories if needed.
  • path: string — Target file or directory path relative to the workspace root directory
  • content: string — Data content to write to the file (text or base64 encoded binary)
  • encoding?: string — Encoding of the provided content string: 'utf-8' for plain text, 'base64' for raw binary bytes
JSON passthrough from provider API—
local_fs_list_directory
List directory contents with file metadata (name, size, modification timestamp, and type).
  • path: string — Target file or directory path relative to the workspace root directory
  • recursive?: boolean — When true, traverses subdirectories recursively to list all nested files
  • pattern?: string — Glob filter pattern to restrict returned entries (e.g. '*.ts' or '**/*.json')
JSON passthrough from provider API—
local_fs_delete
Delete a file or directory from the workspace. Recursive deletion of directories requires recursive: true.
  • path: string — Target file or directory path relative to the workspace root directory
  • recursive?: boolean — When true, deletes a non-empty directory and all its contents recursively
JSON passthrough from provider API—
local_fs_mkdir
Create a new directory and any missing parent directories in the workspace path.
  • path: string — Target file or directory path relative to the workspace root directory
JSON passthrough from provider API—
local_fs_move
Move or rename a file or directory from source path to destination path within the workspace.
  • source: string — Source file or directory path relative to workspace root
  • destination: string — Destination file or directory path relative to workspace root
JSON passthrough from provider API—
local_fs_copy
Copy a file or directory from source to destination path within the workspace.
  • source: string — Source file or directory path to copy
  • destination: string — Target destination path for the copy
JSON passthrough from provider API—
local_fs_search_files
Search for files within the workspace matching a glob pattern (e.g. '**/*.tsx').
  • pattern: string — Glob search pattern (e.g. '*.test.ts' or 'src/**/*.css')
  • path?: string — Directory path within workspace to start searching from (defaults to workspace root)
JSON passthrough from provider API—
local_fs_stat
Get detailed filesystem metadata for a file or directory including byte size, permissions, and ISO timestamps.
  • path: string — Target file or directory path relative to the workspace root directory
JSON passthrough from provider API—
local_fs_grep
Search file contents across the workspace using regular expressions. Returns matching lines and line numbers.
  • pattern: string — Regular expression pattern to search for inside file contents
  • path?: string — Root path or directory to search within (defaults to workspace root)
  • filePattern?: string — Optional glob filter to restrict searched file types (e.g. '*.ts')
JSON passthrough from provider API—
local_fs_watch
Take a directory snapshot and compare with previous snapshot to report modified, created, or deleted files.
  • path: string — Target file or directory path relative to the workspace root directory
JSON passthrough from provider API—
local_fs_diff
Compare two files line by line and generate a unified diff showing changes between them.
  • pathA: string — Relative path to the first file (original base)
  • pathB: string — Relative path to the second file (modified target)
JSON passthrough from provider API—
local_fs_patch
Apply a search-and-replace text patch to an existing file in the workspace.
  • path: string — Target file or directory path relative to the workspace root directory
  • search: string — Exact target substring to find and replace within the file
  • replace: string — Replacement text to substitute in place of the found search target
JSON passthrough from provider API—
local_fs_patch_file
Apply multiple sequential search-and-replace text patches to a file atomically.
  • path: string — Target file or directory path relative to the workspace root directory
  • patches: array<object> — List of search-and-replace string pairs to apply in order
JSON passthrough from provider API—
local_shell_run
Execute an allow-listed command in the workspace directory. Restricted by security sandbox policies.
  • command: string — Command executable name (must be present in allowlist, e.g. 'git', 'bun', 'cargo')
  • args?: array<string> — Array of command-line argument strings passed to the executable
  • timeout?: number — Maximum execution timeout duration in milliseconds (default 30000)
JSON passthrough from provider API—
local_clipboard_read
Read the current textual contents from the system clipboard.
noneJSON passthrough from provider API—
local_clipboard_write
Write text content directly to the system clipboard.
  • text: string — Plain text string to place onto the system clipboard
JSON passthrough from provider API—
local_notification_send
Display a native desktop notification to the user.
  • title: string — Notification header title text
  • message: string — Notification body message text
JSON passthrough from provider API—
local_os_info
Get operating system details including platform, CPU architecture, hostname, and memory metrics.
noneJSON passthrough from provider API—

productivity (10)

Asana asana · 12 tools · Organization credentials

Asana project management integration for tasks, projects, workspaces, and team collaboration. Supports custom fields, timelines, and portfolio tracking.

Transport: cloud + local (stdio) · Required credentials: ASANA_ACCESS_TOKEN

ToolParametersReturnsUser Scoping
asana_list_tasks
List tasks in a project or assigned to a user
  • projectGid?: string — Project GID to list tasks from
  • assignee?: string — User GID or "me" to filter by assignee (requires workspace)
  • workspace?: string — Workspace GID (required when filtering by assignee)
  • completedSince?: string — ISO 8601 date — only return tasks completed since this date. Use "now" to show only incomplete tasks.
JSON passthrough from provider APIparam: `assignee` (caller-asserted)
asana_get_task
Get detailed information about a task
  • taskGid: string — Unique Asana global ID (GID) of the task (e.g. '1200000000000001')
JSON passthrough from provider API—
asana_create_task
Create a new task in a project
  • name: string — Display title or summary line for the new task
  • projectGid?: string — Project GID to add the task to
  • assignee?: string — User GID or "me" to assign the task to
  • dueOn?: string — Due date in YYYY-MM-DD format
  • notes?: string — Task description (plain text)
  • tags?: array<string> — Tag GIDs to apply
JSON passthrough from provider APIparam: `assignee` (caller-asserted)
asana_update_task
Update an existing Asana task's name, completion status, due date, notes, or assignee. Unspecified fields remain unchanged. Returns the updated task object.
  • taskGid: string — Unique Asana global ID (GID) of the task (e.g. '1200000000000001')
  • name?: string — New task name
  • completed?: boolean — Mark task as complete or incomplete
  • dueOn?: string — Due date in YYYY-MM-DD format
  • notes?: string — Task description (plain text)
  • assignee?: string — User GID or "me"
JSON passthrough from provider APIparam: `assignee` (caller-asserted)
asana_add_comment
Add a comment (story) to a task
  • taskGid: string — Unique Asana global ID (GID) of the task (e.g. '1200000000000001')
  • text: string — Comment text
JSON passthrough from provider API—
asana_list_projects
List projects in a workspace
  • workspace: string — Workspace GID
  • archived?: boolean — Filter by archived status
JSON passthrough from provider API—
asana_get_project
Get detailed information about a project
  • projectGid: string — Unique Asana global ID (GID) of the project (e.g. '1200000000000002')
JSON passthrough from provider API—
asana_create_project
Create a new project in a workspace
  • workspace: string — Workspace GID
  • name: string — Project name
  • notes?: string — Project description
  • color?: string — Project color (e.g. "light-green", "dark-blue")
  • defaultView?: string — Default project view
JSON passthrough from provider API—
asana_list_workspaces
List all workspaces the authenticated user belongs to
noneJSON passthrough from provider API—
asana_search_tasks
Search tasks in a workspace
  • workspace: string — Workspace GID
  • text?: string — Text to search for
  • assignee?: string — User GID or "me"
  • projectGid?: string — Restrict search to a project
  • completed?: boolean — Filter by completion status
JSON passthrough from provider APIparam: `assignee` (caller-asserted)
asana_list_sections
List sections in a project
  • projectGid: string — Unique Asana global ID (GID) of the project (e.g. '1200000000000002')
JSON passthrough from provider API—
asana_move_task_to_section
Move a task to a specific section within a project
  • sectionGid: string — Section GID to move the task into
  • taskGid: string — Task GID to move
JSON passthrough from provider API—
Google Calendar google-calendar · 10 tools · Per-user OAuth / credentials

Google Calendar via delegated Google OAuth (users/me). Connect from Desktop — not a shared workspace calendar.

Transport: cloud + local (stdio) · Required credentials: GOOGLE_CLIENT_ID, GOOGLE_CLIENT_SECRET, GOOGLE_ACCESS_TOKEN

ToolParametersReturnsUser Scoping
gcal_list_events
List events from a Google Calendar with optional time range and search query
  • calendarId?: string — Calendar ID (default: "primary")
  • timeMin?: string — Lower bound (RFC3339, e.g. "2026-03-20T00:00:00Z")
  • timeMax?: string — Upper bound (RFC3339)
  • maxResults?: number — Maximum number of events to return
  • query?: string — Free text search query
JSON passthrough from provider API—
gcal_get_event
Get detailed information about a specific calendar event
  • calendarId?: string — Calendar ID (default: "primary")
  • eventId: string — Unique alphanumeric Google Calendar event identifier (e.g. 'abc123xyz')
JSON passthrough from provider API—
gcal_create_event
Create a new event on a Google Calendar
  • calendarId?: string — Calendar ID (default: "primary")
  • summary: string — Event title
  • description?: string — Event description
  • startDateTime: string — Start time (RFC3339, e.g. "2026-03-21T10:00:00-07:00")
  • endDateTime: string — End time (RFC3339)
  • location?: string — Event location
  • attendees?: array<string> — Array of attendee email addresses
  • reminders?: object — Reminder overrides (e.g. { useDefault: false, overrides: [{ method: "popup", minutes: 10 }] })
JSON passthrough from provider APIparam: `attendees` (caller-asserted)
gcal_update_event
Update an existing calendar event
  • calendarId?: string — Calendar ID (default: "primary")
  • eventId: string — Unique alphanumeric Google Calendar event identifier (e.g. 'abc123xyz')
  • summary?: string — Updated event title
  • description?: string — Updated event description
  • startDateTime?: string — Updated start time (RFC3339)
  • endDateTime?: string — Updated end time (RFC3339)
  • location?: string — Updated event location
JSON passthrough from provider API—
gcal_delete_event
Delete a calendar event permanently from Google Calendar and notify attendees.
  • calendarId?: string — Calendar ID (default: "primary")
  • eventId: string — Event ID to delete
JSON passthrough from provider API—
gcal_list_calendars
List all calendars the user has access to
noneJSON passthrough from provider API—
gcal_quick_add
Quickly create an event from a natural language string (e.g. "Lunch with Bob tomorrow at noon")
  • calendarId?: string — Calendar ID (default: "primary")
  • text: string — Natural language event description
JSON passthrough from provider API—
gcal_freebusy
Check free/busy status for one or more calendars over a time range
  • timeMin: string — Start of time range (RFC3339)
  • timeMax: string — End of time range (RFC3339)
  • calendars?: array<string> — Array of calendar IDs to check (default: ["primary"])
JSON passthrough from provider API—
gcal_rsvp
Respond to an event invitation (accept, decline, or tentative)
  • calendarId?: string — Calendar ID (default: "primary")
  • eventId: string — Event ID to respond to
  • response: string — RSVP response
JSON passthrough from provider API—
gcal_list_upcoming
List the next N upcoming events from now
  • count?: number — Number of upcoming events to return
  • calendarId?: string — Calendar ID (default: "primary")
JSON passthrough from provider API—
Jira jira · 14 tools · Organization credentials

Atlassian Jira integration for issue tracking, sprint management, backlog grooming, and project reporting. Supports Jira Cloud and Data Center.

Transport: cloud + local (stdio) · Required credentials: JIRA_BASE_URL, JIRA_EMAIL, JIRA_API_TOKEN

ToolParametersReturnsUser Scoping
jira_search_issues
Search for Jira issues using JQL (Jira Query Language)
  • jql: string — JQL query string (e.g. "project = PROJ AND status = Open")
  • maxResults?: number — Maximum results to return (default 50, max 100)
  • fields?: string — Comma-separated field names to include (e.g. "summary,status,assignee")
JSON passthrough from provider API—
jira_get_issue
Get detailed information about a specific Jira issue
  • issueKey: string — Issue key (e.g. PROJ-123)
JSON passthrough from provider API—
jira_create_issue
Create a new issue, bug report, or task in Jira with project key and summary.
  • projectKey: string — Project key (e.g. PROJ)
  • summary: string — Issue summary/title
  • description?: string — Issue description (plain text, converted to ADF)
  • issueType: string — Issue type categorization: 'Task' for operational work, 'Bug' for defects, 'Story' for feature requirements, 'Epic' for high-level initiatives
  • priority?: string — Priority name (e.g. High, Medium, Low)
  • assignee?: string — Assignee account ID
  • labels?: array<string> — Labels to apply
JSON passthrough from provider APIparam: `assignee` (caller-asserted)
jira_update_issue
Update fields on an existing Jira issue
  • issueKey: string — Issue key (e.g. PROJ-123)
  • summary?: string — New summary
  • description?: string — New description (plain text, converted to ADF)
  • priority?: string — New priority name
  • assignee?: string — New assignee account ID
  • labels?: array<string> — New labels (replaces existing)
JSON passthrough from provider APIparam: `assignee` (caller-asserted)
jira_transition_issue
Move an issue to a new status via a workflow transition
  • issueKey: string — Issue key (e.g. PROJ-123)
  • transitionId?: string — Transition ID (use jira_get_transitions to find available IDs)
  • transitionName?: string — Transition name (alternative to transitionId — matched case-insensitively)
JSON passthrough from provider API—
jira_add_comment
Add a comment to a Jira issue
  • issueKey: string — Issue key (e.g. PROJ-123)
  • body: string — Comment body (plain text, converted to ADF)
JSON passthrough from provider API—
jira_list_projects
List all accessible Jira projects
noneJSON passthrough from provider API—
jira_get_project
Get detailed information about a Jira project
  • projectKey: string — Project key (e.g. PROJ)
JSON passthrough from provider API—
jira_list_sprints
List sprints for a specific board
  • boardId: number — Numeric Jira agile board identifier (e.g. 42)
  • state?: string — Filter by sprint state
JSON passthrough from provider API—
jira_get_sprint
Get details for a specific sprint
  • sprintId: number — Numeric Jira sprint identifier (e.g. 101)
JSON passthrough from provider API—
jira_list_boards
List all accessible Jira boards
noneJSON passthrough from provider API—
jira_assign_issue
Assign a Jira issue to a user
  • issueKey: string — Issue key (e.g. PROJ-123)
  • accountId: string — Atlassian account ID of the assignee
JSON passthrough from provider API—
jira_get_transitions
Get available workflow transitions for an issue
  • issueKey: string — Issue key (e.g. PROJ-123)
JSON passthrough from provider API—
jira_link_issues
Create a link between two Jira issues
  • inwardIssue: string — Issue key for the inward side (e.g. PROJ-123)
  • outwardIssue: string — Issue key for the outward side (e.g. PROJ-456)
  • linkType: string — Link type name (e.g. "Blocks", "Relates", "Duplicates", "Clones")
JSON passthrough from provider API—
Nunc Markitdown markitdown · 2 tools · Built-in (no external credentials)

Convert documents (PDF, DOCX, XLSX, PPTX, images) to clean Markdown. Powered by Microsoft Markitdown CLI with OCR support.

Transport: cloud + local (stdio)

ToolParametersReturnsUser Scoping
markitdown_convert
Convert a document to markdown. Accepts a local file path or URL. Supports PDF, DOCX, XLSX, PPTX, HTML, images, CSV, JSON, XML, EPub, and more.
  • source: string — File path or URL to convert. For local files, provide an absolute path. For remote content, provide a full URL (https://...).
JSON passthrough from provider API—
markitdown_convert_base64
Convert a base64-encoded document to markdown. Useful when file content is passed inline rather than by path. Supports PDF, DOCX, XLSX, PPTX, HTML, images, CSV, JSON, XML, EPub, and more.
  • data: string — Base64-encoded file content.
  • filename: string — Original filename with extension (e.g. "report.pdf", "slides.pptx"). The extension is used to determine the file type.
JSON passthrough from provider API—
Nunc Audio Generation nunc-audio-gen · 1 tools · Built-in (no external credentials)

Generate spoken audio from text prompts through Nunc's governed speech synthesis pipeline with authentication and audit logging.

Transport: cloud + local (stdio)

ToolParametersReturnsUser Scoping
audio_speech_generate
Convert text to high-quality spoken audio (speech synthesis). Returns the generated audio file URL.
  • text: string — The text content to synthesize into speech
  • voice?: string — Optional voice preset identifier.
JSON passthrough from provider API—
Nunc Image Generation nunc-image-gen · 3 tools · Built-in (no external credentials)

Generate and edit images from text prompts through Nunc's governed image pipeline with authentication and audit logging.

Transport: cloud + local (stdio)

ToolParametersReturnsUser Scoping
image_generate
Generate an image from a text prompt. Returns a URL to the generated image.
  • prompt: string — Detailed text description of the image to generate
  • size?: number — Image size in pixels (square). Default: 1024
  • style?: string — Visual style for the generated image. Default: realistic
  • negative_prompt?: string — Things to exclude from the generated image (e.g. "blurry, low quality")
JSON passthrough from provider API—
image_edit
Edit an existing image based on a text instruction. Accepts a base64-encoded image or URL plus an edit prompt.
  • image: string — Source image as a base64-encoded string or a URL
  • prompt: string — Text instruction describing the edit to apply
  • mask?: string — Optional base64-encoded mask image (white = edit region, black = keep)
JSON passthrough from provider API—
image_describe
Analyze an image and return a text description of its contents.
  • image: string — Image to describe, as a base64-encoded string or a URL
JSON passthrough from provider API—
Nunc Maps & Geospatial nunc-map · 3 tools · Built-in (no external credentials)

Geocoding, location lookup, and static map rendering. Routes through the gateway for auth and audit.

Transport: cloud + local (stdio)

ToolParametersReturnsUser Scoping
map_geocode
Convert a physical address or place name into geographic coordinates (latitude and longitude).
  • address: string — Full street address, city, or landmark name (e.g. '1600 Amphitheatre Pkwy, Mountain View, CA')
JSON passthrough from provider API—
map_search_places
Search for specific places, points of interest, or businesses nearby or in a specific region. Returns GeoJSON FeatureCollection.
  • query: string — Search query (e.g. 'restaurants', 'parks', 'coffee')
  • latitude: number — Latitude center coordinate
  • longitude: number — Longitude center coordinate
  • radius?: number — Search radius in meters. Default: 1000
JSON passthrough from provider API—
map_calculate_route
Calculate a route between start and destination coordinates. Returns route details and GeoJSON LineString route path.
  • startLatitude: number — Starting point latitude
  • startLongitude: number — Starting point longitude
  • endLatitude: number — Destination latitude
  • endLongitude: number — Destination longitude
  • mode?: string — Mode of transport (e.g., 'driving', 'walking', 'bicycling'). Default: 'driving'
JSON passthrough from provider API—
Nunc Video Generation nunc-video-gen · 3 tools · Built-in (no external credentials)

Generate a video from a text prompt, animate a still image, or edit an existing clip through Nunc's governed media pipeline.

Transport: cloud + local (stdio)

ToolParametersReturnsUser Scoping
video_generate
Generate a short video clip from a text prompt. Returns a URL to the clip. Billed per second (480p default).
  • prompt: string — Detailed text description of the video clip to generate (e.g. 'Golden retriever puppy running on a sunny beach')
  • duration?: number — Length in seconds, from 1 to 15. Default: 5
  • aspect_ratio?: string — Frame shape. Default: 16:9
  • resolution?: string — 480p (default) or 720p. 720p costs more per second.
JSON passthrough from provider API—
video_animate_image
Animate a still image into a short video clip. Pass an http(s) image URL or base64. Returns a URL to the clip.
  • imageUrl: string — http(s) URL or base64 of the still image to animate. Private and loopback addresses are refused.
  • prompt?: string — How to animate the image (e.g. 'make the waves crash gently')
  • duration?: number — Length in seconds, from 1 to 15. Default: 5
  • aspect_ratio?: string — Frame shape. Default: 16:9. Image-to-video may follow the source image instead.
  • resolution?: string — 480p (default) or 720p.
JSON passthrough from provider API—
video_edit
Edit an existing clip from a text instruction. Pass an http(s) video URL or base64 and an instruction. The edited clip keeps the source timing. Returns a URL.
  • videoUrl: string — http(s) URL or base64 of the video to edit. Private and loopback addresses are refused.
  • prompt: string — Instruction on how to edit the video (e.g. 'add a vintage color filter')
JSON passthrough from provider API—
Nunc Web nunc-web · 9 tools · Built-in (no external credentials)

Managed headless web search, dynamic page browsing, screenshot capture, and PDF document generation.

Transport: cloud + local (stdio)

ToolParametersReturnsUser Scoping
web_search
Search the web without third-party API keys and return ranked organic results (title, url, snippet).
  • query: string — Search query string.
  • count?: number — Max results to return (1-20). Defaults to 5.
JSON passthrough from provider API—
web_browse
Navigate to a URL using serverless Chromium, evaluate dynamic client JavaScript (React, Next.js, etc.), and extract clean Markdown.
  • url: string — The absolute HTTP/HTTPS URL to navigate to.
  • extract?: string — Extraction format. Defaults to 'markdown'.
  • wait_for?: string — Optional CSS selector to wait for before extracting content.
  • sessionId?: string — Optional browser session ID for multi-turn interactions.
JSON passthrough from provider API—
web_pdf
Render a web page URL or HTML template into a pixel-perfect PDF document saved to tenant storage.
  • url?: string — URL to render into PDF.
  • html?: string — Raw HTML string to render into PDF.
  • format?: string — Paper size format. Defaults to 'Letter'.
  • landscape?: boolean — Print in landscape orientation. Defaults to false.
  • sessionId?: string — Optional browser session ID.
JSON passthrough from provider API—
web_screenshot
Capture a visual screenshot PNG of a web page or specific element saved to tenant storage.
  • url: string — URL to capture.
  • selector?: string — Optional CSS selector of a specific element to capture.
  • fullPage?: boolean — Capture full scrollable page. Defaults to false.
  • sessionId?: string — Optional browser session ID.
JSON passthrough from provider API—
web_click
Click an element on the page using a CSS selector.
  • selector: string — CSS selector of the element to click.
  • sessionId?: string — Browser session ID to run the click within.
JSON passthrough from provider API—
web_fill
Type text into an input field or textarea using a CSS selector.
  • selector: string — CSS selector of the input field.
  • value: string — The text value to type into the field.
  • sessionId?: string — Browser session ID to run the fill within.
JSON passthrough from provider API—
web_evaluate
Evaluate a JavaScript expression or script in the active page context.
  • script: string — JavaScript code or expression to evaluate.
  • sessionId?: string — Browser session ID to evaluate within.
JSON passthrough from provider API—
web_scroll
Scroll the active page or a scrollable element container up, down, to top, or to bottom.
  • direction?: string — Scroll direction. Defaults to 'down'.
  • amount?: number — Number of pixels to scroll. Defaults to 800.
  • selector?: string — Optional CSS selector of a scrollable container element.
  • sessionId?: string — Browser session ID to scroll within.
JSON passthrough from provider API—
web_hover
Hover mouse cursor over an element using a CSS selector to trigger hover states or dropdown menus.
  • selector: string — CSS selector of the element to hover over.
  • sessionId?: string — Browser session ID to hover within.
JSON passthrough from provider API—
SurveyMonkey surveymonkey · 45 tools · Organization credentials

First-party SurveyMonkey REST API v3 integration for surveys, pages, questions, aggregated survey results and rollups, respondent answers, trends, benchmarks, collectors, email invitations, contacts, teams, and webhooks.

Transport: cloud + local (stdio) · Required credentials: SURVEYMONKEY_ACCESS_TOKEN, SURVEYMONKEY_BASE_URL

SurveyMonkey REST API v3 (workspace/account-tier Bearer token, US/EU/CA regional base URL support). 45 tools across Survey Results & Analytics (composite results joining `/details` + `/rollups` + `/responses/bulk` open-ended verbatim answers, question/survey rollups with computed percentages & statistical summaries, time-series `/trends`, `/benchmark`, and bulk/single respondent answers with automatic ID-to-label resolution), Survey/Page/Question authoring, Templates, Folders, Question Bank, Collectors/Messages/Send/Stats, Contacts/Contact Lists, Webhooks, and Groups/Workgroups.

ToolParametersReturnsUser Scoping
surveymonkey_get_me
Retrieve the authenticated SurveyMonkey user profile, account type, email, username, and granted OAuth scopes (GET /v3/users/me).
noneJSON passthrough from provider API—
surveymonkey_list_surveys
List surveys in the SurveyMonkey account with optional title search, folder filter, modification date range, and sorting (GET /v3/surveys). Always includes response_count, question_count, date_modified, and analyze_url by default.
  • title?: string — Search surveys by title substring.
  • folder_id?: string — Filter to surveys inside a specific folder ID.
  • sort_by?: string — Field to sort by (default: date_modified).
  • sort_order?: string — Sort direction (default: DESC).
  • start_modified_at?: string — Surveys modified after YYYY-MM-DDTHH:MM:SS (no timezone offset).
  • end_modified_at?: string — Surveys modified before YYYY-MM-DDTHH:MM:SS (no timezone offset).
  • include?: string — Comma-separated fields/filters (default: 'response_count,date_created,date_modified,language,question_count,analyze_url,preview'). Can also include 'shared_with', 'shared_by', or 'owned'.
  • page?: number — Page number (1-based, default 1).
  • per_page?: number — Surveys per page (default 50, max 1000).
  • max_pages?: number — Number of pages to auto-paginate in a single call (default 1, max 25).
JSON passthrough from provider API—
surveymonkey_get_survey
Retrieve metadata for a specific survey including response_count, page_count, question_count, custom_variables, and analyze/collect/edit URLs (GET /v3/surveys/{survey_id}).
  • survey_id: string — Unique SurveyMonkey survey ID.
JSON passthrough from provider API—
surveymonkey_get_survey_details
Retrieve the full nested design of a survey including all pages, questions, headings, and answer choices/rows/columns (GET /v3/surveys/{survey_id}/details).
  • survey_id: string — Unique SurveyMonkey survey ID.
JSON passthrough from provider API—
surveymonkey_create_survey
Create a new survey — either blank, cloned from an existing survey (from_survey_id) or template (from_template_id), or populated with initial pages and questions (POST /v3/surveys).
  • title: string — Title of the new survey.
  • nickname?: string — Internal nickname for the survey.
  • language?: string — ISO language code (e.g. 'en', 'es', 'fr').
  • folder_id?: string — Folder ID to place the survey in.
  • from_survey_id?: string — Existing survey ID to clone structure from.
  • from_template_id?: string — SurveyMonkey template ID to clone from.
  • custom_variables?: object — Key-value map of custom variable names to labels for tracking query parameters.
  • pages?: array<object> — Optional array of page objects (with title, description, questions[]) to create inline.
JSON passthrough from provider API—
surveymonkey_update_survey
Update a survey's title, nickname, language, folder_id, custom_variables, or buttons_text (PATCH /v3/surveys/{survey_id}).
  • survey_id: string — Survey ID to update.
  • title?: string — Updated survey title.
  • nickname?: string — Updated internal nickname.
  • language?: string — Updated language code.
  • folder_id?: string — Move survey to folder ID.
  • custom_variables?: object — Updated custom variables map.
  • buttons_text?: object — Button labels ({ next_button, prev_button, done_button, exit_button }).
JSON passthrough from provider API—
surveymonkey_delete_survey
Delete a survey by ID (DELETE /v3/surveys/{survey_id}).
  • survey_id: string — Survey ID to delete.
JSON passthrough from provider API—
surveymonkey_list_folders
List survey folders used to organize surveys (GET /v3/survey_folders).
  • page?: number — Page number (default 1).
  • per_page?: number — Folders per page (default 50).
JSON passthrough from provider API—
surveymonkey_create_folder
Create a new folder for organizing surveys (POST /v3/survey_folders).
  • title: string — Folder title (max 32 characters).
JSON passthrough from provider API—
surveymonkey_list_templates
List official SurveyMonkey survey templates that can be passed as from_template_id to surveymonkey_create_survey (GET /v3/survey_templates).
  • category?: string — Filter templates by category name.
  • language?: string — Filter templates by language code (e.g. 'en').
  • show_only_available_to_current_user?: boolean — Only return templates unlocked on the current account plan.
  • page?: number — Page number (default 1).
  • per_page?: number — Templates per page (default 50).
JSON passthrough from provider APIparam: `show_only_available_to_current_user` (caller-asserted)
surveymonkey_list_question_bank
Search SurveyMonkey's certified Question Bank for pre-written questions (GET /v3/question_bank/questions).
  • search?: string — Keyword search across Question Bank questions.
  • category?: string — Question bank category filter.
  • language?: string — Language code (default 'en').
  • page?: number — Page number (default 1).
  • per_page?: number — Results per page (default 50).
JSON passthrough from provider API—
surveymonkey_list_pages
List all pages in a survey (GET /v3/surveys/{survey_id}/pages).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • page?: number — Page number (default 1).
  • per_page?: number — Results per page (default 50).
JSON passthrough from provider API—
surveymonkey_create_page
Add a new page to a survey (POST /v3/surveys/{survey_id}/pages).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • title?: string — Page heading title.
  • description?: string — Page subtitle / instructions.
  • position?: number — 1-based position of the page in the survey.
JSON passthrough from provider API—
surveymonkey_update_page
Update a survey page's title, description, or position (PATCH /v3/surveys/{survey_id}/pages/{page_id}).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • page_id: string — Page ID to update.
  • title?: string — Updated page title.
  • description?: string — Updated page description.
  • position?: number — Updated page position.
JSON passthrough from provider API—
surveymonkey_delete_page
Delete a page from a survey (DELETE /v3/surveys/{survey_id}/pages/{page_id}).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • page_id: string — Page ID to delete.
JSON passthrough from provider API—
surveymonkey_list_questions
List questions on a specific survey page (GET /v3/surveys/{survey_id}/pages/{page_id}/questions).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • page_id: string — Unique SurveyMonkey survey page identifier (from surveymonkey_list_pages).
  • page?: number — Page number (default 1).
  • per_page?: number — Questions per page (default 50).
JSON passthrough from provider API—
surveymonkey_get_question
Get full details of a specific question including headings, family, subtype, validation, and answer choices (GET /v3/surveys/{survey_id}/pages/{page_id}/questions/{question_id}).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • page_id: string — Unique SurveyMonkey survey page identifier (from surveymonkey_list_pages).
  • question_id: string — Unique SurveyMonkey question identifier (from surveymonkey_list_questions).
JSON passthrough from provider API—
surveymonkey_create_question
Create a question on a survey page (POST /v3/surveys/{survey_id}/pages/{page_id}/questions). Supports single_choice, multiple_choice, open_ended, matrix, demographic, datetime, presentation, etc.
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • page_id: string — Unique SurveyMonkey survey page identifier (from surveymonkey_list_pages).
  • heading?: string — Question prompt text (convenience field; automatically wrapped into headings: [{ heading }] if headings array is omitted).
  • headings?: array<object> — SurveyMonkey headings array (e.g. [{ heading: 'How satisfied are you?' }]).
  • family: string — Question family.
  • subtype: string — Question subtype (e.g. 'vertical', 'horiz', 'menu' for choice; 'essay', 'single', 'multi', 'numerical' for open_ended; 'rating', 'single', 'ranking', 'menu' for matrix; ' descriptive_text' for presentation).
  • answers?: object — Answers definition object (e.g. { choices: [{ text: 'Very Satisfied' }, { text: 'Satisfied' }] } or { rows: [...], choices: [...] } for matrix). Omit for open_ended essay/single.
  • choices?: array<string> — Convenience array of answer choice labels (e.g. ['Very Satisfied', 'Satisfied', 'Neutral']) — automatically converted into answers.choices.
  • rows?: array<string> — Convenience array of matrix row labels (e.g. ['Speed', 'Quality']) — automatically converted into answers.rows.
  • required?: object — Optional required validation object (e.g. { text: 'Please answer this question', type: 'all', amount: '1' }).
  • position?: number — 1-based position on the page.
JSON passthrough from provider API—
surveymonkey_update_question
Update an existing question's headings, answers, position, or validation settings (PATCH /v3/surveys/{survey_id}/pages/{page_id}/questions/{question_id}).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • page_id: string — Unique SurveyMonkey survey page identifier (from surveymonkey_list_pages).
  • question_id: string — Question ID to update.
  • heading?: string — Updated question heading text.
  • headings?: array<object> — Updated headings array.
  • answers?: object — Updated answers object (choices, rows, cols, other).
  • choices?: array<string> — Convenience array of updated answer choice labels — automatically converted into answers.choices.
  • rows?: array<string> — Convenience array of updated matrix row labels — automatically converted into answers.rows.
  • required?: object — Updated required validation configuration.
  • position?: number — Updated position on the page.
JSON passthrough from provider API—
surveymonkey_delete_question
Delete a question from a survey page (DELETE /v3/surveys/{survey_id}/pages/{page_id}/questions/{question_id}).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • page_id: string — Unique SurveyMonkey survey page identifier (from surveymonkey_list_pages).
  • question_id: string — Question ID to delete.
JSON passthrough from provider API—
surveymonkey_get_survey_results
Retrieve the complete, human-readable Survey Results report for a survey. Combines /surveys/{id}/details and /surveys/{id}/rollups to return every question's exact text, answer choice counts, computed percentages, completion rates, and statistical summaries (mean, median, min, max, std). When include_open_ended=true (default), also fetches recent responses to attach verbatim text/essay answers under each open-ended question.
Primary single-call survey results tool for Claude & Nunc Chat — resolves all opaque question/choice/row IDs into human-readable labels, computes percentages & completion rates, and attaches recent open-ended text answers.
  • survey_id: string — Unique SurveyMonkey survey ID.
  • include_open_ended?: boolean — Also fetch recent bulk responses and attach verbatim text answers to open-ended/other questions (default: true).
  • open_ended_limit?: number — Max recent responses to sample for open-ended text answers (default: 50, max: 100).
  • collector_ids?: string — Comma-separated collector IDs to filter results to specific campaigns/links.
  • status?: string — Filter by response status: 'completed', 'partial', 'overquota', 'disqualified' (comma-separated allowed).
  • start_created_at?: string — Include responses created after this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • end_created_at?: string — Include responses created before this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • start_modified_at?: string — Include responses last modified after this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • end_modified_at?: string — Include responses last modified before this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • email?: string — Filter to responses from a specific email address (email collectors only).
  • first_name?: string — Filter to responses matching respondent first name (email collectors only).
  • last_name?: string — Filter to responses matching respondent last name (email collectors only).
  • ip?: string — Filter to responses from a specific IP address.
  • custom?: string — Filter by custom metadata value associated with respondents.
  • total_time_min?: number — Minimum time spent on the survey.
  • total_time_max?: number — Maximum time spent on the survey.
  • total_time_units?: string — Unit for total_time_min / total_time_max.
Composite report joining `/surveys/{id}/details`, `/surveys/{id}/rollups`, and `/surveys/{id}/responses/bulk` verbatim answers
Key fields: survey_id, title, response_count, questions[].question_heading, questions[].summary[].choices[], questions[].sample_text_responses[]
param: `email` (caller-asserted)
surveymonkey_get_survey_rollup
Retrieve aggregated question rollups (counts, percentages, and stats) for all questions in a survey (GET /v3/surveys/{survey_id}/rollups). Automatically resolves question and choice IDs to human-readable text when resolve_labels=true (default).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • resolve_labels?: boolean — Join /surveys/{id}/details to attach question_heading, choice text, and percentages (default: true).
  • page?: number — Page number (default 1).
  • per_page?: number — Questions per page (default 50).
  • collector_ids?: string — Comma-separated collector IDs to filter results to specific campaigns/links.
  • status?: string — Filter by response status: 'completed', 'partial', 'overquota', 'disqualified' (comma-separated allowed).
  • start_created_at?: string — Include responses created after this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • end_created_at?: string — Include responses created before this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • start_modified_at?: string — Include responses last modified after this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • end_modified_at?: string — Include responses last modified before this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • email?: string — Filter to responses from a specific email address (email collectors only).
  • first_name?: string — Filter to responses matching respondent first name (email collectors only).
  • last_name?: string — Filter to responses matching respondent last name (email collectors only).
  • ip?: string — Filter to responses from a specific IP address.
  • custom?: string — Filter by custom metadata value associated with respondents.
  • total_time_min?: number — Minimum time spent on the survey.
  • total_time_max?: number — Maximum time spent on the survey.
  • total_time_units?: string — Unit for total_time_min / total_time_max.
JSON passthrough from provider APIparam: `email` (caller-asserted)
surveymonkey_get_question_rollup
Retrieve aggregated answer counts, percentages, and statistical summary (mean, median, min, max, std) for a single survey question (GET /v3/surveys/{survey_id}/pages/{page_id}/questions/{question_id}/rollups).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • page_id: string — Unique SurveyMonkey survey page identifier (from surveymonkey_list_pages).
  • question_id: string — Unique SurveyMonkey question identifier (from surveymonkey_list_questions).
  • resolve_labels?: boolean — Resolve choice/row IDs to human-readable labels and compute percentages (default: true).
  • collector_ids?: string — Comma-separated collector IDs to filter results to specific campaigns/links.
  • status?: string — Filter by response status: 'completed', 'partial', 'overquota', 'disqualified' (comma-separated allowed).
  • start_created_at?: string — Include responses created after this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • end_created_at?: string — Include responses created before this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • start_modified_at?: string — Include responses last modified after this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • end_modified_at?: string — Include responses last modified before this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • email?: string — Filter to responses from a specific email address (email collectors only).
  • first_name?: string — Filter to responses matching respondent first name (email collectors only).
  • last_name?: string — Filter to responses matching respondent last name (email collectors only).
  • ip?: string — Filter to responses from a specific IP address.
  • custom?: string — Filter by custom metadata value associated with respondents.
  • total_time_min?: number — Minimum time spent on the survey.
  • total_time_max?: number — Maximum time spent on the survey.
  • total_time_units?: string — Unit for total_time_min / total_time_max.
JSON passthrough from provider APIparam: `email` (caller-asserted)
surveymonkey_get_survey_trends
Retrieve time-series response trends across all questions in a survey broken down by hour, day, week, month, quarter, or year (GET /v3/surveys/{survey_id}/trends). Automatically resolves question and choice labels when resolve_labels=true (default).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • trend_by?: string — Time bucket granularity (default: 'day' or 'month').
  • resolve_labels?: boolean — Join /surveys/{id}/details to attach question and choice labels (default: true).
  • first_respondent?: string — Limit to responses after this respondent ID.
  • last_respondent?: string — Limit to responses before this respondent ID.
  • page?: number — Page number (default 1).
  • per_page?: number — Questions per page (default 50).
  • collector_ids?: string — Comma-separated collector IDs to filter results to specific campaigns/links.
  • status?: string — Filter by response status: 'completed', 'partial', 'overquota', 'disqualified' (comma-separated allowed).
  • start_created_at?: string — Include responses created after this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • end_created_at?: string — Include responses created before this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • start_modified_at?: string — Include responses last modified after this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • end_modified_at?: string — Include responses last modified before this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • email?: string — Filter to responses from a specific email address (email collectors only).
  • first_name?: string — Filter to responses matching respondent first name (email collectors only).
  • last_name?: string — Filter to responses matching respondent last name (email collectors only).
  • ip?: string — Filter to responses from a specific IP address.
  • custom?: string — Filter by custom metadata value associated with respondents.
  • total_time_min?: number — Minimum time spent on the survey.
  • total_time_max?: number — Maximum time spent on the survey.
  • total_time_units?: string — Unit for total_time_min / total_time_max.
JSON passthrough from provider APIparam: `email` (caller-asserted)
surveymonkey_list_responses
Retrieve full respondent submissions in bulk for a survey or specific collector (GET /v3/surveys/{survey_id}/responses/bulk or /v3/collectors/{collector_id}/responses/bulk). When resolve_labels=true (default), resolves all question IDs and choice/row/column IDs into human-readable text and attaches a flattened `resolved_answers` list on each response.
Supports `email` filter for per-user respondent lookup; automatically resolves choice/row IDs to text labels (`resolve_labels=true` by default).
  • survey_id: string — Survey ID (required for survey-wide responses and label resolution).
  • collector_id?: string — Optional collector ID to pull responses from a single collector.
  • resolve_labels?: boolean — Resolve question and choice IDs to human-readable text (default: true).
  • sort_by?: string — Sort field (SurveyMonkey supports 'date_modified').
  • sort_order?: string — Sort order (default: DESC for newest responses first).
  • page?: number — Page number (default 1).
  • per_page?: number — Responses per page (default 50, max 100).
  • max_pages?: number — Number of pages to auto-paginate in a single call (default 1, max 25 — up to 2,500 bulk responses).
  • collector_ids?: string — Comma-separated collector IDs to filter results to specific campaigns/links.
  • status?: string — Filter by response status: 'completed', 'partial', 'overquota', 'disqualified' (comma-separated allowed).
  • start_created_at?: string — Include responses created after this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • end_created_at?: string — Include responses created before this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • start_modified_at?: string — Include responses last modified after this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • end_modified_at?: string — Include responses last modified before this ISO timestamp (YYYY-MM-DDTHH:MM:SS).
  • email?: string — Filter to responses from a specific email address (email collectors only).
  • first_name?: string — Filter to responses matching respondent first name (email collectors only).
  • last_name?: string — Filter to responses matching respondent last name (email collectors only).
  • ip?: string — Filter to responses from a specific IP address.
  • custom?: string — Filter by custom metadata value associated with respondents.
  • total_time_min?: number — Minimum time spent on the survey.
  • total_time_max?: number — Maximum time spent on the survey.
  • total_time_units?: string — Unit for total_time_min / total_time_max.
JSON passthrough from provider APIparam: `email` (caller-asserted)
surveymonkey_get_response
Retrieve a single respondent's complete survey response with all page/question answers and human-readable labels (GET /v3/surveys/{survey_id}/responses/{response_id}/details).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • response_id: string — Unique SurveyMonkey respondent submission identifier (from surveymonkey_list_responses).
  • resolve_labels?: boolean — Resolve question and choice IDs into human-readable labels (default: true).
JSON passthrough from provider API—
surveymonkey_create_response
Submit or import a response for a survey collector (POST /v3/collectors/{collector_id}/responses).
  • collector_id: string — Collector ID to record the response under.
  • pages: array<object> — Array of page answer objects: [{ id: page_id, questions: [{ id: question_id, answers: [{ choice_id } | { text } | { row_id, choice_id }] }] }].
  • response_status?: string — Response status (default: 'completed').
  • custom_variables?: object — Custom variable key-value pairs for the respondent.
  • custom_value?: string — Custom identifier string for the response.
  • recipient_id?: string — Optional email collector recipient ID.
  • ip_address?: string — Respondent IP address (optional).
JSON passthrough from provider API—
surveymonkey_delete_response
Delete a specific survey response by ID (DELETE /v3/surveys/{survey_id}/responses/{response_id}).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • response_id: string — Response ID to delete.
JSON passthrough from provider API—
surveymonkey_get_question_benchmark
Retrieve industry benchmark comparison statistics (percentiles, average, respondent count) for a Question Bank question in a survey (GET /v3/surveys/{survey_id}/pages/{page_id}/questions/{question_id}/benchmark).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • page_id: string — Unique SurveyMonkey survey page identifier (from surveymonkey_list_pages).
  • question_id: string — Unique SurveyMonkey question identifier (from surveymonkey_list_questions).
  • bundle_id?: string — Optional benchmark bundle ID (e.g. 'nps_education') to compare against.
JSON passthrough from provider API—
surveymonkey_list_collectors
List distribution collectors (weblinks, email campaigns, SMS, popups) for a survey along with their URLs, status, and response counts (GET /v3/surveys/{survey_id}/collectors).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • name?: string — Filter collectors by name substring.
  • sort_by?: string — Field name used to order the returned records.
  • sort_order?: string — Sort direction.
  • page?: number — Page number (default 1).
  • per_page?: number — Collectors per page (default 50).
  • max_pages?: number — Number of pages to auto-paginate in a single call (default 1, max 25).
JSON passthrough from provider API—
surveymonkey_get_collector
Retrieve full configuration and status for a collector, optionally including its response stats (GET /v3/collectors/{collector_id} and /v3/collectors/{collector_id}/stats).
  • collector_id: string — Unique SurveyMonkey collector identifier (from surveymonkey_list_collectors).
  • include_stats?: boolean — Also fetch /collectors/{id}/stats and attach it as `stats` (default: true).
JSON passthrough from provider API—
surveymonkey_create_collector
Create a new survey collector such as a shareable weblink or email invitation collector (POST /v3/surveys/{survey_id}/collectors).
  • survey_id: string — Unique SurveyMonkey survey identifier (from surveymonkey_list_surveys).
  • type: string — Collector type ('weblink' returns a shareable URL).
  • name?: string — Collector name.
  • thank_you_message?: string — Custom thank-you message displayed on completion.
  • redirect_url?: string — URL to redirect respondents to after completion.
  • close_date?: string — Auto-close timestamp (ISO-8601).
  • anonymous_type?: string — Respondent anonymity mode.
  • allow_multiple_responses?: boolean — Allow a respondent to take the survey more than once.
  • response_limit?: number — Maximum number of responses before closing.
JSON passthrough from provider API—
surveymonkey_update_collector
Update a collector's status ('open' or 'closed'), name, close_date, redirect_url, or anonymity settings (PATCH /v3/collectors/{collector_id}).
  • collector_id: string — Unique SurveyMonkey collector identifier (from surveymonkey_list_collectors).
  • name?: string — Updated collector name.
  • status?: string — Open or close the collector.
  • close_date?: string — Updated close date.
  • thank_you_message?: string — Updated thank-you message.
  • redirect_url?: string — Updated redirect URL.
  • allow_multiple_responses?: boolean — Allow multiple responses per respondent.
  • response_limit?: number — Updated response cap.
JSON passthrough from provider API—
surveymonkey_delete_collector
Delete a collector by ID (DELETE /v3/collectors/{collector_id}).
  • collector_id: string — Collector ID to delete.
JSON passthrough from provider API—
surveymonkey_list_collector_messages
List email or SMS messages configured on an invitation collector (GET /v3/collectors/{collector_id}/messages).
  • collector_id: string — Unique SurveyMonkey collector identifier (from surveymonkey_list_collectors).
  • page?: number — Page number (default 1).
  • per_page?: number — Messages per page (default 50).
JSON passthrough from provider API—
surveymonkey_create_collector_message
Create an invitation, reminder, or thank_you message on an email/SMS collector, and optionally attach recipients or contact lists (POST /v3/collectors/{collector_id}/messages).
  • collector_id: string — Email collector ID.
  • type: string — Email collector message category (invite, reminder, or thank_you).
  • recipient_status?: string — Target audience filter (required for 'reminder' and 'thank_you' messages).
  • subject?: string — Email subject line.
  • body_text?: string — Plain text body (must include [SurveyLink], [OptOutLink], and [FooterLink] placeholders if required by SurveyMonkey).
  • body_html?: string — HTML email body.
  • is_branding_enabled?: boolean — Include SurveyMonkey branding.
  • contacts?: array<object> — Optional array of recipients ({ email, first_name?, last_name?, custom_fields? }) to add immediately to this message via /recipients/bulk.
  • contact_list_ids?: array<string> — Optional array of contact_list IDs to add as recipients to this message.
JSON passthrough from provider APIparam: `contacts` (caller-asserted)
surveymonkey_send_collector_message
Send or schedule a prepared collector email/SMS message (POST /v3/collectors/{collector_id}/messages/{message_id}/send).
  • collector_id: string — Unique SurveyMonkey collector identifier (from surveymonkey_list_collectors).
  • message_id: string — Message ID to send.
  • scheduled_date?: string — Optional ISO-8601 timestamp to schedule delivery (omit to send immediately).
JSON passthrough from provider API—
surveymonkey_list_contact_lists
List contact lists in the SurveyMonkey address book (GET /v3/contact_lists).
  • page?: number — Page number (default 1).
  • per_page?: number — Contact lists per page (default 50).
JSON passthrough from provider API—
surveymonkey_create_contact_list
Create a new contact list in the SurveyMonkey address book (POST /v3/contact_lists).
  • name: string — Name of the contact list.
JSON passthrough from provider API—
surveymonkey_list_contacts
List or search contacts globally or within a specific contact_list_id (GET /v3/contacts or /v3/contact_lists/{contact_list_id}/contacts).
  • contact_list_id?: string — Optional contact list ID to list contacts from (omit for global address book).
  • status?: string — Filter contacts by delivery status.
  • search_by?: string — Field to search against.
  • search?: string — Search term matching search_by.
  • sort_by?: string — Field name used to order the returned records.
  • sort_order?: string — Sort direction for ordering returned records (ASC or DESC).
  • page?: number — Page number (default 1).
  • per_page?: number — Contacts per page (default 50).
  • max_pages?: number — Number of pages to auto-paginate in a single call (default 1, max 25).
JSON passthrough from provider API—
surveymonkey_create_contacts_bulk
Create one or more contacts in bulk, either globally or inside a specific contact_list_id (POST /v3/contacts/bulk or /v3/contact_lists/{contact_list_id}/contacts/bulk).
  • contact_list_id?: string — Optional contact list ID to add the contacts into (omit for global contacts).
  • contacts: array<object> — Array of contact objects: [{ email, first_name?, last_name?, custom_fields?: { '1'?: string, ... } }].
  • update_existing?: boolean — Overwrite existing contacts matching by email (default: true).
JSON passthrough from provider APIparam: `contacts` (caller-asserted)
surveymonkey_list_webhooks
List registered webhooks on the SurveyMonkey account (GET /v3/webhooks).
  • page?: number — Page number (default 1).
  • per_page?: number — Webhooks per page (default 50).
JSON passthrough from provider API—
surveymonkey_create_webhook
Create a webhook subscription to receive real-time event callbacks such as response_completed (POST /v3/webhooks).
  • name: string — Descriptive webhook name.
  • event_type: string — Event type to subscribe to.
  • object_type?: string — Whether object_ids refer to surveys or collectors (omit for survey_created).
  • object_ids?: array<string> — Array of survey IDs or collector IDs to watch.
  • subscription_url: string — HTTPS callback URL that receives webhook POSTs.
JSON passthrough from provider API—
surveymonkey_delete_webhook
Delete a webhook subscription by ID (DELETE /v3/webhooks/{webhook_id}).
  • webhook_id: string — Webhook ID to delete.
JSON passthrough from provider API—
surveymonkey_list_groups_and_workgroups
Retrieve SurveyMonkey team groups and workgroups for the organization (GET /v3/groups and /v3/workgroups).
  • include_workgroups?: boolean — Also fetch /v3/workgroups and return both groups and workgroups (default: true).
JSON passthrough from provider API—

storage (2)

Google Drive MCP google-drive · runtime-discovered · Per-user OAuth / credentials

Federates Google's official Drive MCP (drivemcp.googleapis.com) through Nunc /mcp. Each Nunc user Connects Drive as themselves — a second Connect, not Gmail. Writes are catalogued but not auto-allowed. Developer Preview.

Transport: cloud

Tools are discovered dynamically at runtime via tools/list once connected.

OneDrive onedrive · 10 tools · Per-user OAuth / credentials

Microsoft OneDrive personal drive via delegated Microsoft Graph (/me/drive). People-lane user OAuth — not application permissions.

Transport: cloud + local (stdio) · Required credentials: MICROSOFT_TENANT_ID, MICROSOFT_CLIENT_ID, MICROSOFT_CLIENT_SECRET, MICROSOFT_ACCESS_TOKEN

ToolParametersReturnsUser Scoping
onedrive_list_files
List files and folders in a directory. Use root or specify a folder ID or drive path.
  • itemId?: string — Folder item ID or relative drive path (e.g. "Documents/Reports"; omit for root)
  • top?: number — Number of items to return
  • select?: string — Comma-separated fields to select
JSON passthrough from provider API—
onedrive_get_file
Get metadata for a specific file or folder
  • itemId: string — Microsoft Graph item ID or relative drive path (e.g. "Documents/report.pdf")
JSON passthrough from provider API—
onedrive_download_file
Download file content (returns base64-encoded content)
  • itemId: string — File item ID or relative drive path (e.g. "Documents/report.pdf")
JSON passthrough from provider API—
onedrive_upload_file
Upload a file to OneDrive by path (Graph simple-upload limit 4MB). Content should be base64-encoded.
  • path: string — Destination path (e.g. "Documents/report.pdf")
  • content: string — Base64-encoded file content
  • contentType?: string — MIME type of the file
JSON passthrough from provider API—
onedrive_create_folder
Create a new folder in OneDrive under the root or within a parent folder.
  • parentId?: string — Parent folder item ID or relative drive path (e.g. "Documents/Projects"; omit for root)
  • name: string — Folder name
JSON passthrough from provider API—
onedrive_delete_item
Permanently delete a file or folder from OneDrive by item ID or path.
  • itemId: string — Item ID or relative drive path to delete
JSON passthrough from provider API—
onedrive_move_item
Move a file or folder to a new location
  • itemId: string — Item ID or relative drive path to move
  • parentId: string — Destination parent folder item ID or relative drive path
  • newName?: string — Optional new name for the item
JSON passthrough from provider API—
onedrive_search
Search for files and folders by name or content
  • query: string — Search query
  • top?: number — Number of results to return
JSON passthrough from provider API—
onedrive_share
Create a sharing link for a file or folder
  • itemId: string — Item ID or relative drive path to share
  • type?: string — Sharing link permission type: 'view' grants read-only access, 'edit' grants write access, 'embed' generates embeddable iframe HTML
  • scope?: string — Link scope
JSON passthrough from provider API—
onedrive_get_recent
Get recently accessed files
noneJSON passthrough from provider API—

Built-in Platform Tools (41)

Memory, workflow runs, isolated code execution, and automation tools included on /mcp for every tenant without enabling an external connector. Tenant and user identity are bound directly from your OAuth session or bearer key.

ToolParametersReturnsUser Scoping
nunc_search_memory
Hybrid search across the tenant's memory graph (FTS + vector ANN with RRF fusion). Returns top-K knowledge nodes with relevance scores.
  • query: string — Search query text.
  • limit?: integer — Max results to return. Default 10.
  • node_type?: string — Filter by node type (e.g., 'profile', 'context', 'knowledge', 'skill').
  • min_confidence?: number — Minimum confidence score (0-1). Default 0.0.
  • include_historical?: boolean — Include historical facts (expired or future-dated). Default false.
  • as_of?: string — Point-in-time timestamp (ISO string or epoch ms). Defaults to current time.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_recall_context
Recover conversation history that was compacted away. When the context shows a <CONTEXT_SUMMARY> pointing at a segment, use action='list' to see the compacted segments, then action='fetch' with a segment number to read those turns verbatim.
  • action?: string — 'list' shows available segments (default); 'fetch' returns one segment's full transcript.
  • segment?: integer — Segment number to fetch (from action='list').
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_recall_strategies
Recall reusable reasoning strategies (ReasoningBank) relevant to a task — patterns distilled from this tenant's past successful and failed workflow run trajectories. Consult before planning open-ended work to repeat what worked and avoid what didn't.
  • query: string — The task or goal to find relevant strategies for (natural language).
  • limit?: number — Max strategies to return (1-20). Default 5.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_ingest
Persist a fact, decision, skill, or piece of context into the tenant's memory graph. Returns the new node id.
  • content: string — Text content of the knowledge node.
  • node_type?: string — One of 'profile', 'context', 'knowledge', 'skill', 'fact', 'decision', 'geospatial'.
  • category?: string — Domain category (e.g., 'finance', 'donor', 'governance', 'KNOWLEDGE').
  • confidence?: number — Confidence score (0-1). Default 0.8.
  • tags?: array<string> — Free-form tags.
  • geojson?: object — Optional GeoJSON FeatureCollection to persist as a geospatial knowledge node via persistGeoArtifact.
  • metadata?: object — Optional metadata object attached to the knowledge node.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_query_dataset
Compute over an already-spilled tool result (`datasets/<tenantId>/…` or `/api/datasets/…`) or an explicitly uploaded catalog file dataset (`ds_…`). Do NOT invent `dataset_id` values. To query connected SQL databases or organization tables (e.g. members, rsvps, events, accounts, companies, payments), ALWAYS use `nunc_query_datasource` instead.
  • dataset_id: string — An existing spilled dataset key (datasets/<tenantId>/…), spilled dataset url (/api/datasets/…), or uploaded file dataset id (ds_…). Never guess or invent this ID.
  • query?: object — Structured query object. Omit `aggregate` for a filtered/selected page of rows; include `aggregate` (+ optional `groupBy`) to aggregate. Shape: { where?: [{col, op:'='|'!='|'>'|'>='|'<'|'<='|'contains', value}], groupBy?: [col], aggregate?: [{col, fn:'sum'|'count'|'avg'|'min'|'max', as?}], select?: [col], orderBy?: [{col, dir:'asc'|'desc'}], limit?: number }.
  • where?: array<object> — Optional top-level filter clauses: [{ col: string, op: '='|'!='|'>'|'>='|'<'|'<='|'contains', value: string|number }].
  • groupBy?: array<string> — Optional top-level columns to group by.
  • aggregate?: array<object> — Optional top-level aggregations: [{ col: string, fn: 'sum'|'count'|'avg'|'min'|'max', as?: string }].
  • select?: array<string> — Optional top-level columns to select.
  • orderBy?: array<object> — Optional top-level sort order: [{ col: string, dir?: 'asc'|'desc' }].
  • limit?: number — Optional max rows to return (default 200).
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_query_graphql
Execute a semantic GraphQL query against the tenant's dynamic GraphQL Yoga schema. Preferred for querying SQL tables with inferred relation counts, aggregations, federated MCP/API tools, Clairic forecasts, R2 lakehouse datasets, and business anomalies. Available Query fields: `<tableName>(where: JSON, orderBy: String, limit: Int, offset: Int) { <col> <childTable>_count <childTable>_<milestone>_count mcpEnrich(tool: String!, matchKey: String) }`, `<tableName>_aggregate(groupBy: [String!]!, sum: [String!], avg: [String!], where: JSON, orderBy: String, limit: Int, forecastHorizon: Int) { key count <childTable>_count <childTable>_<milestone>_count conversionRatePct sums avgs forecast }`, `databaseTables { name connectionId columns }`, `mcpTools { name connector operationType readOnly description }`, `callMcpTool(name: String!, args: JSON)`, `forecastSeries(values: [Float!]!, horizon: Int, timestamps: [String!])`, `extractStructured(text: String!, schema: JSON!)`, `datasets { id name rowCount columns { name type } }`, `datasetQuery(id: String!, where: JSON, groupBy: [String!], aggregate: [JSON!], limit: Int) { columns rows rowCount }`, `anomalies { id metricName observedValue baselineValue variancePercent severity source description suggestedAction }`, `latestDigest { summary activeWorkflowsCount avgLatencyMs }`.
  • query: string — A valid GraphQL query string against the tenant's dynamic GraphQL Yoga schema (discover table & column names from your schema context or via `databaseTables` / `mcpTools`).
  • variables?: object — Optional GraphQL variables map.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_query_datasource
Run read-only SQL (SELECT or WITH…SELECT) against a connected SQL database (D1, Postgres, DuckDB, Snowflake, etc.) for organization data such as members, rsvps, events, companies, opportunities, payments, or telemetry. If `connection_id` is omitted and the tenant has one connected database, it is selected automatically.
  • connection_id?: string — The data-source connection id (e.g. 'd1:ecw-app'). Optional when only one SQL data source is connected.
  • sql: string — A single read-only SQL statement (SELECT or WITH…SELECT). No writes/DDL.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_check_safety
Run the 5-pillar Guardian pipeline on text — sanitize + 17-category PII/secrets + prompt injection & SQL/GraphQL query guard + content/compliance policy (Clef @cf/cloudflare/clef-flash) + egress control. Returns the verdict.
  • content: string — Text to inspect.
  • stage?: string — Optional: 'input' or 'output' (affects which Guardian stages run). Default 'input'.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_video_analyze
Answer a question about an ingested video from its timestamped transcript and visual chunks (video-ingest pipeline). Videos must first be uploaded via POST /api/video/upload. Cites timestamps.
  • query: string — The question to answer about the video.
  • video?: string — Optional filename / job-id substring to scope to one video. Omit to search all ingested videos.
  • limit?: integer — Max transcript/visual chunks to read (1-40).
  • model?: string — Optional model override (allow-list validated, e.g. '@cf/google/gemma-4-26b-a4b-it').
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_update_tasks
Read or update the live markdown checklist (tasks.md) for the active conversation or goal. Supports '- [ ]' (todo), '- [/]' (in_progress), '- [x]' (done), and '- [!]' (blocked). Omit tasks_md to read the current checklist.
  • tasks_md?: string — Optional markdown checklist lines (e.g. '- [ ] Step 1\n- [x] Step 2') to reconcile into tasks.
  • parent_id?: string — Optional parent goal task ID to scope the checklist to.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_load_skill
Load the full SKILL.md instructions, output formats, and examples for an installed skill by slug or name on demand.
  • slug: string — The skill slug or name from AVAILABLE SKILLS INDEX.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_invoke_subagent
Delegate parallel specialist sub-agents (`subagents`) or run a multi-step Goal Loop checklist (`tasks_md`).
  • goal: string — Natural-language description of the work to be done.
  • tasks_md?: string — Optional markdown checklist ('- [ ] step 1\n- [ ] step 2') to execute sequentially in a Goal Loop.
  • priority?: string — 'normal' (default), 'high', 'low'. Influences execution order.
  • agency?: string — Slug or ID of an Agency from the agency catalog (e.g. 'finance', 'engineering', 'sales'). Automatically resolves conductor and specialist sub-agents.
  • agent_id?: string — Optional agent persona ID to execute the Goal Loop or sub-agent task under.
  • subagents?: array<object> — Array of parallel AI sub-agent specialist specs ({ name, role, taskPrompt, tools, depends, maxTurns }). Fanned out in parallel waves into a synthesis step.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_cancel_workflow_run
Terminate a running or awaiting-approval Workflow Run (`AgentWorkflow`) for this tenant.
  • run_id: string — The Workflow Run ID to cancel.
  • reason?: string — Optional human-readable cancellation reason.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_approve_workflow_step
Approve a Workflow Run or step parked at `awaiting_approval` or `awaiting_review`, resuming `AgentWorkflow`.
  • run_id: string — The Workflow Run ID to approve.
  • step_id?: string — Optional step ID when approving a specific parked checklist step.
  • reason?: string — Optional approval note.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_reject_workflow_step
Reject a Workflow Run or step parked at `awaiting_approval` or `awaiting_review`, cancelling execution.
  • run_id: string — The Workflow Run ID to reject.
  • step_id?: string — Optional step ID being rejected.
  • reason?: string — Optional rejection reason.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_open_gadget
Create or return a durable first-class canvas for this conversation (one of the 4 First-Class Canvases: 'dashboard', 'presentation', 'spreadsheet', 'document'), or look one up by id. Always pass the structured content in `body`: for 'dashboard' pass {title, subtitle?, kpis:[{label,value,delta?,trend?}], charts:[{title, type:'bar'|'line'|'area'|'pie'|'map'|'diagram'|'event', data:[...], geojson?, code?, events?}]}; for 'presentation' pass {title, subtitle?, slides:[{id, title, subtitle?, layout, bullets?, metrics?, chart?, notes?}]}; for 'spreadsheet' pass {title?, columns?, rows:[...], scenarios?:[{name,field,min,max,step,defaultValue}]}; for 'document' pass {title?, markdown:'...', diagrams?:[{title,code}]}. Returns {id, format, title, body}. Do not reprint the full canvas in chat.
  • id?: string — Existing gadget id. When set, returns that gadget (tenant-scoped) and ignores format. A body still updates the row.
  • format?: string — Required when creating. One kept format.
  • title?: string — Optional display title.
  • body?: object — Body to store (e.g. {html:'...'}, {markdown:'...'}, {layout:[], widgets:[]}, {type:'bar', data:[]}, {type:'FeatureCollection', deckLayer:'hexagon', features:[]}). Written on create and when a row for this conversation and format already exists.
  • html?: string — Convenience HTML markup when format is 'html'.
  • markdown?: string — Convenience Markdown content when format is 'document'.
  • conversation_id?: string — Conversation to bind. Defaults to the current conversation.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_patch_gadget
Apply a structured in-place patch to a gadget. Do NOT reprint the whole artifact. Document: {op:'set_markdown',markdown} or {op:'replace',target,replacement}. Presentation: {op:'set_slot',slideId|slideIndex,slotId|role,text,value} or {op:'set_element',slideIndex,elementId,content}. HTML/chart/dashboard/spreadsheet/map: {op:'set_body',body:{...}} (or pass body/html/markdown directly). Returns {id, format, title, body}.
  • id?: string — Gadget id from nunc_open_gadget. Optional when format matches this conversation's gadget.
  • gadget_id?: string — Alias for id.
  • format?: string — Used with the current conversation to resolve the gadget when id is omitted.
  • patch?: object — Structured patch object with an `op` field (set_markdown, replace, set_slot, set_element, set_body) and `body` / `markdown` / `html`.
  • body?: object — Optional replacement body object when updating the gadget body directly.
  • html?: string — Optional replacement HTML string for html gadgets.
  • markdown?: string — Optional replacement Markdown string for document gadgets.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_create_trigger
Register an event-driven automation (a trigger) — fire a skill, workflow, or agent when a matching event arrives. Use for 'when X happens, do Y' phrasing. External-write triggers are created DISARMED and must be armed in the Automations dashboard; read-only triggers self-arm.
  • name: string — Short human-readable name for the automation.
  • event_source: string — The inbound event source that fires this trigger.
  • condition?: object — Optional match condition (field/operator/value predicate). Omit to fire on every event from the source.
  • target_kind: string — What the trigger runs when it fires.
  • target_ref: string — Reference to the target — a skill slug, a saved workflow name/id, or an agent id.
  • read_only?: boolean — True (default) if the automation only reads. Set false for external-write automations — those are created DISARMED and require dashboard approval.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_update_trigger
Update an existing event-driven automation (a trigger) by id — rename it, change its event_source/condition/target, or toggle read_only/armed. Only the fields you pass are changed. Write-safe: setting read_only=false (turning it into an external-write automation) forces armed=false, so it can never be silently armed via an update — re-arm it in the Automations dashboard. Returns isError if no trigger matches the id.
  • trigger_id: string — The id of the trigger to update.
  • name?: string — New human-readable name for the automation.
  • event_source?: string — New inbound event source that fires this trigger.
  • condition?: object — New match condition (field/operator/value predicate). Pass an empty object to fire on every event from the source.
  • target_kind?: string — What the trigger runs when it fires.
  • target_ref?: string — Reference to the target — a skill slug, a saved workflow name/id, or an agent id.
  • read_only?: boolean — Set false to mark this as an external-write automation — doing so forces armed=false (re-arm via the dashboard).
  • armed?: boolean — Whether the trigger fires on matching events. Ignored (forced false) when read_only is being set to false.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_schedule
Register a time-driven automation (a scheduled task). Use for 'every / each week / at 9am / on Monday' phrasing. Provide exactly one of cron (recurring, 5-field) or fire_at (one-shot ISO timestamp). Pass saved_workflow to schedule a reusable workflow; otherwise the prompt runs as a chat-mode task.
  • title: string — Short title for the scheduled task.
  • prompt?: string — What the task should do when it runs. Optional when saved_workflow is given (a prompt is synthesized from the workflow).
  • cron?: string — 5-field cron expression for a recurring schedule (e.g. '0 9 * * 1' = Mondays 9am). Mutually exclusive with fire_at.
  • fire_at?: string — ISO 8601 timestamp for a one-shot run. Mutually exclusive with cron.
  • timezone?: string — IANA timezone (e.g. 'America/New_York'). Default UTC.
  • saved_workflow?: string — Optional saved workflow name or id to run on schedule. When set, the task runs in act mode.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_search_findings
Surface what the background-agent fleet has observed and learned for this tenant: distilled operational rules (learnings, tenant-wide) plus recent agent findings. Use to consult prior failures/corrections before acting.
  • limit?: integer — Max findings to return (1-100). Default 20.
  • min_severity?: string — Filter findings to this severity or higher: 'info' | 'warning' | 'error'.
  • include_learnings?: boolean — Include distilled SYSTEM rules (default true). Set false for findings only.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_execute_code
Execute custom code (JavaScript/TypeScript or Python), routed by 'target': cloud runs in an isolated Cloudflare Sandbox rooted at `/workspace` (the scale lane for parallel/heavy jobs), local runs on the user's machine via a connected sidecar. Never the gateway worker.
  • code: string — The source code to execute.
  • language: string — The execution language ('javascript' | 'typescript' | 'python').
  • stdin?: string — Optional standard input to pass to the script execution.
  • timeout?: integer — Optional execution timeout limit in seconds.
  • target?: string — Where to run. 'auto' (default): local when a workspace sidecar is connected, else cloud. 'local': force on-machine (sidecar). 'cloud': force the Cloudflare Sandbox — the scale lane for massive-parallel/heavy jobs. Local supports javascript/python without stdin; anything else runs cloud.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_preview_app
Start a long-running web server in the sandbox and return a public live preview URL. Use for apps that serve on a port (Flask/FastAPI/Streamlit/Gradio/Vite) — NOT for static HTML (emit a <nunc:web-preview> block instead) or one-shot scripts (use nunc_execute_code). Write the project with nunc_execute_code first; this runs the server against those same files. Returns { url, port }. After it returns, present the url in a ```url:preview fenced code block so the desktop opens a live preview pane.
  • command: string — Shell command that starts the server bound to `port` (e.g. 'python3 -m http.server 8000 --directory /workspace', 'streamlit run /workspace/app.py --server.port 8501', 'npm run dev').
  • port: integer — The port the server listens on inside the sandbox.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_stage_file
Pull a laptop-uploaded file into the sandbox filesystem so code can use it (cross-lane data bridge). The user's desktop uploads a local file via POST /api/sandbox/stage, which returns an R2 `key`; pass that key here with a `dest_path` to land it in the sandbox. Use for local inputs (CSV/GIS/CAD/images) that nunc_execute_code then processes. Runs in the same user session as nunc_execute_code.
  • key: string — R2 staging key returned by POST /api/sandbox/stage (the uploaded local file).
  • dest_path: string — Absolute path under `/workspace` inside the sandbox to write the file to (e.g. '/workspace/data.csv').
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_harvest_file
Push a file the sandbox produced back out to R2 staging so it can be delivered to the user's laptop / Files (cross-lane data bridge). Returns an R2 `key` the desktop downloads via GET /api/sandbox/stage?key=. Use after nunc_execute_code generates a deliverable (report PDF, rendered map, exported dataset).
  • src_path: string — Absolute path under `/workspace` of the file inside the sandbox to harvest (e.g. '/workspace/report.pdf').
  • name?: string — Optional friendly filename for the delivered artifact (defaults to the basename of src_path).
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_clarify
Ask the user a multiple-choice or text question when clarification/feedback is needed.
  • question: string — The question to present to the user.
  • choices?: array<string> — Optional array of choices (strings) for multiple choice selection.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_request_resource
Ask the user to introduce a SaaS resource (GitHub repo, Salesforce record, Cvent event) into this conversation. Does not grant access itself — returns needs_introduction so the UI can prompt Add resource.
  • connector?: string — Connector slug to request (e.g. github, salesforce, cvent).
  • resource_url?: string — URL of the resource to introduce, if known.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_proposed_actions
List business actions the platform proposed from learned subsystem activity, awaiting approval. Approve a proposal via POST /api/workflows/runs/:id/approve; dismiss via POST /api/workflows/runs/:id/reject.
  • limit?: integer — Max proposals to return (1-100). Default 20.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_workspace_patch
Apply search-and-replace patches to a file on the local filesystem. Supports loose whitespace/indentation matching to prevent failures.
  • path: string — File path (relative to local workspace root)
  • patches: array<object> — List of search-and-replace patches to apply in sequence
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_workspace_run_command
Run an allow-listed command or script locally in the workspace. The allowlist is Console → Compute → Local Commands (0.4.1+ sidecars) or each device's NUNC_SHELL_ALLOWLIST. Pass the binary name in `command` and arguments in `args` — not a shell string.
  • command: string — The allow-listed command to run (e.g., git, ls).
  • args?: array<string> — Arguments to pass to the command.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_dispatch_webhook
Dispatch a governed HTTP POST/PUT/PATCH payload to an external webhook or API endpoint. Protected against SSRF; only public HTTPS endpoints are allowed.
  • url: string — Target HTTPS webhook or API URL.
  • payload?: object — JSON data payload to transmit.
  • method?: string — HTTP method: POST (default), PUT, or PATCH.
  • headers?: object — Optional custom HTTP headers.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_webhook_dispatch
Alias for nunc_dispatch_webhook. Dispatch a governed HTTP payload to an external webhook endpoint.
  • url: string — Target HTTPS webhook or API URL.
  • payload?: object — JSON data payload to transmit.
  • method?: string — HTTP method: POST (default), PUT, or PATCH.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_forecast_series
Deterministic 64-step time-series forecasting via Clairic (Cloudflare System One). Generates forward projections with calibrated confidence bounds, trend classification, and mathematical summary at $0 output token cost.
  • values: array<number> — Historical numerical values to project forward.
  • horizon?: integer — Number of steps forward to forecast (1 to 64, default 12).
  • timestamps?: array<string> — Optional ISO timestamps matching the historical values.
  • confidence_interval?: number — Confidence interval level (default 0.95).
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_extract_structured
Extract structured fields from unformatted text, tables, or JSON dumps using Clairic. Returns strongly typed values (text, number, datetime, list, estimate) with zero hallucination.
  • text: string — Unstructured text or tabular payload to extract from.
  • schema: object — Field definitions mapping field names to types (text, number, datetime, list, estimate).
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_analytics_schema
Introspect the per-user RBAC-scoped semantic data graph, SQL tables, connected MCP tools, and Lakehouse datasets available for the authenticated user and tenant. Supports optional System 1 Clef domain pruning when a natural-language query is provided.
  • entity?: string — Optional entity or table name to inspect (e.g. 'members', 'datasets', 'anomalies')
  • query?: string — Optional natural-language question so System 1 Clef (@cf/cloudflare/clef-flash) can rank and prune the user's authorized MCP tools and tables to the most relevant domain
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_analytics_query
Execute a federated GraphQL Yoga query against Nunc's semantic layer (SQL tables, `_aggregate`, connected MCP tools, Lakehouse datasets, and Clairic `forecastSeries`/`extractStructured`) with per-user RBAC/RLS, edge caching, Guardian PII masking, and optional System 1/2 sufficiency & synthesis.
  • query: string — The GraphQL query to execute against the GraphQL Yoga Gateway
  • variables?: object — Optional query variables as key-value pairs
  • prompt?: string — Optional natural-language question for System 1 Clef sufficiency verification and optional System 2 synthesis
  • effort?: string — Optional System 2 reasoning effort ('fast' -> Gemma 4, 'smart' -> Claude/Grok) when synthesizing an answer for 'prompt'
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_analytics_insights
Query Nunc's observer learning loop and business anomaly sweeps across Data Sources and MCP connections. Supports optional System 1 (Clef) + System 2 ('fast' | 'smart') executive synthesis when 'prompt' is provided.
  • category?: string — Filter by category (e.g. operational, financial, performance, pipeline, data_quality)
  • minConfidence?: number — Minimum confidence score between 0.0 and 1.0 (default 0.5)
  • severity?: string — Filter anomalies by severity: INFO, WARNING, or CRITICAL
  • source?: string — Filter anomalies by source: lakehouse, bigquery, snowflake, stripe, salesforce, mcp
  • since?: string — ISO datetime string to filter anomalies detected on or after this timestamp
  • prompt?: string — Optional analytical question to synthesize over the retrieved anomalies and learning signals via System 1 (Clef) + System 2 reasoning
  • effort?: string — Optional System 2 reasoning effort ('fast' | 'balanced' | 'smart')
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_analytics_propose_action
Stage a governed write-back action (e.g. updating accounting records, Salesforce, re-running a sync, or spawning a remediation workflow). Generates an approval diff parked at a Guardian approval gate.
  • via: string — Target connector or system slug (e.g. lakehouse, salesforce, intacct)
  • actionType: string — Type of action (e.g. UPDATE, SYNC, REMEDIATE)
  • targetEntity: string — Target entity ID or name
  • payload: string — JSON string of changes to apply
  • reason: string — Business justification for this action
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_analytics_create_report
Generate an interactive report session in Nunc Data Studio Web with System 1 Clef canvas format selection ('dashboard' | 'spreadsheet' | 'deck' | 'report') and System 1 Clairic 64-step time-series forecasting.
  • title: string — Report title
  • summary: string — Executive summary of findings
  • format?: string — Optional canvas format. When omitted, System 1 Clef automatically selects the optimal format from title and summary.
  • charts?: array<object> — Array of chart configurations or ECharts JSON specs
  • tables?: array<object> — Array of tabular data objects or dataset queries
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_sandbox_eval
Evaluate a single JavaScript expression inside a tenant-isolated Cloudflare Sandbox and return its typed JSON value. Useful for quick calculations, data transforms, or inspecting sandbox state without writing a full script.
  • expression: string — The JavaScript expression to evaluate.
  • sessionId?: string — Optional sandbox session context ID so expression evaluation shares state with prior runs.
JSON (gateway-native)context-enforced (tenant+user from auth)
nunc_sandbox_status
Inspect Nunc Sandbox availability, supported languages, execution limits, and structural isolation guarantees for this deployment.
noneJSON (gateway-native)context-enforced (tenant+user from auth)

Default-deny 3-tier authorization

Every tool call is evaluated against User → Group → Organization policies in Console → Access. Any tool without an explicit allow rule is denied by default, and mutating tools can require human-in-the-loop approval before executing.