mcp-server-outline
Provides tools for interacting with Outline's API, enabling searching, reading, writing documents, managing collections and comments on an Outline knowledge base.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@mcp-server-outline search for onboarding documents in the Engineering collection"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
mcp-server-outline
MCP server for the Outline knowledge base / wiki API. Works against any self-hosted or cloud Outline instance (we run it against our "Playbook" instances).
Tools
Read: outline_search (ranked full-text with context snippets; filter by collection/status/date/author), outline_list_documents (recent/drafts/archived/trash), outline_get_document (info or full Markdown export), outline_list_collections, outline_get_collection (+ document tree), outline_list_comments, outline_list_revisions, outline_whoami, outline_search_capabilities.
Write (opt-in via OUTLINE_ENABLE_WRITES=true): outline_create_document, outline_update_document (replace or append), outline_move_document, outline_archive_document (archive/restore), outline_delete_document (soft-delete to Trash only), outline_create_comment, outline_manage_collection (create/update).
Deliberately not exposed: collections.delete (cascading permanent delete of every document inside), permanent document deletion / documents.empty_trash, and user deletion. Document deletion is always the recoverable soft-delete.
Related MCP server: mcp-outline
Auth
Bearer API token (Outline → Settings → API → Create). The token inherits its creating user's document/collection permissions — create a dedicated service user per instance with the right collection access. Tokens are per-instance.
Configuration
See .env.example. Required: OUTLINE_BASE_URL (the server appends /api), OUTLINE_API_TOKEN.
For the HTTP transport set MCP_HTTP_TOKEN; the bearer token is compared in constant time. API method names are restricted to the collection.method form, so caller input is never placed in the URL path.
Run
npm install
npm run dev # stdio
npm run dev:http # streamable HTTP on :3000/mcp (stateless)
npm testDocker images: ghcr.io/borgels/mcp-server-outline (published on push to main).
Available Tools
16 toolsoutline_archive_documentArchive/Restore Document (Outline)A
Archive a document (reversible) or restore an archived/trashed one. Requires write access.
Inputs: id (required), action (required).
outline_create_commentComment on Document (Outline)A
Add a comment to a document; parentCommentId replies to a thread. Requires write access.
Inputs: text (required), documentId (required), parentCommentId.
outline_create_documentCreate Document (Outline)A
Create a document (Markdown body) in a collection. publish=true makes it visible; parentDocumentId nests it. Requires write access.
Inputs: title (required), collectionId (required), text, publish, +1 more.
outline_delete_documentDelete Document (Outline)A
Move a document to Trash (recoverable via restore). Permanent deletion is not available here. Requires write access.
Inputs: id (required).
outline_get_collectionGet Collection (Outline)ARead-onlyIdempotent
Fetch one collection; includeStructure=true adds its document tree.
Inputs: id (required), includeStructure.
outline_get_documentGet Document (Outline)ARead-onlyIdempotent
Fetch one document by id or slug. asMarkdown=true returns the full Markdown body.
Inputs: id (required), asMarkdown.
outline_list_collectionsList Collections (Outline)CRead-onlyIdempotent
List the collections in the workspace.
Inputs: limit, offset.
outline_list_commentsList Comments (Outline)CRead-onlyIdempotent
List comments on a document.
Inputs: documentId (required), limit, offset.
outline_list_documentsList Documents (Outline)ARead-onlyIdempotent
List documents: recent (default), drafts, archived, or deleted (Trash). Optionally scope to a collection.
Inputs: limit, filter, offset, collectionId.
outline_list_revisionsList Revisions (Outline)CRead-onlyIdempotent
A document’s version history, or one revision with revisionId.
Inputs: documentId (required), limit, offset, revisionId.
outline_manage_collectionCreate/Update Collection (Outline)A
Create a new collection or update one (name/description/permission). Deleting collections is not available (cascades permanently). Requires write access.
Inputs: action (required), id, name, color, +2 more.
outline_move_documentMove Document (Outline)A
Move a document to another collection and/or under another parent document. Requires write access.
Inputs: id (required), collectionId, parentDocumentId.
outline_searchSearch Documents (Outline)ARead-onlyIdempotent
Ranked full-text search with context snippets. Optional filters: collectionId, statusFilter (draft/published/archived), dateFilter (day/week/month/year), userId.
Inputs: query (required), limit, offset, userId, +3 more.
outline_search_capabilitiesSearch Outline CapabilitiesARead-onlyIdempotent
Search the Outline/Playbook MCP capabilities. Use first when deciding which tool to call.
Inputs: limit, query.
outline_update_documentUpdate Document (Outline)A
Update a document title/body. append=true adds text to the end; otherwise text REPLACES the whole body. Requires write access.
Inputs: id (required), text, title, append, +1 more.
outline_whoamiWho Am I (Outline)ARead-onlyIdempotent
Show the user + workspace this Playbook token belongs to.
Inputs: no named parameters.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 16 tools
Most tools have clearly distinct purposes (create, update, delete, move, archive documents; manage collections; list comments/revisions; search). However, 'outline_search_capabilities' searches MCP capabilities rather than documents, which could cause confusion with 'outline_search'. Also, 'outline_archive_document' combines archive and restore actions into one tool, slightly blurring the operation.
All tools follow a consistent 'outline_<verb>_<noun>' pattern (e.g., outline_create_document, outline_list_comments). Verbs are descriptive and nouns match the resource. The pattern is uniform across the entire set, making it predictable for an agent.
16 tools is an appropriate number for a document management server. It covers essential operations on documents, collections, comments, revisions, and search without being overwhelming or sparse.
The tool set covers full CRUD for documents (create, read, update, delete/trash, move, archive) and collections (create, update, list, get). It also includes comments, revisions, search, and user info. Minor gaps: no permanent deletion (only trash), no document publishing workflow beyond publish flag, and no user/group management.
Related MCP Connectors
Read spaces, collections, pages and content; search docs and manage GitBook organization spaces.
Search, read and create Linear issues, projects, teams and cycles.
Search and read your meeting notes, transcripts and action items, and create notes in Reline.
Access and edit your Craft documents.
Related MCP Servers
- FlicenseCqualityNot gradedmaintenanceEnables LLMs to interact with Outline wiki for document management, search, collections, and comments, with optional AI-powered features including RAG-based Q\&A and content summarization.37513 npm-
- FlicenseNot gradedqualityAmaintenanceEnables AI assistants to interact with Outline documentation services, including searching, reading, creating, editing, and archiving documents, managing collections and comments, and finding backlinks.-
- FlicenseNot gradedqualityAmaintenanceEnables AI assistants to interact with Outline documentation services by searching, reading, creating, editing, and archiving documents, managing collections, and working with comments and backlinks.-
- FlicenseNot gradedqualityAmaintenanceEnables AI assistants to interact with Outline documentation services for searching, reading, creating, editing, and archiving documents, managing collections and comments, and retrieving backlinks and table-of-contents data.-