Outline

This MCP server lets you query and manage your Outline wiki from large language models through structured API calls. It adds optional AI-powered features for semantic search, Q&A, and smart insights, while still supporting the standard document CRUD and collaboration operations.

How to use

You connect your MCP client to the Outline-based server using a standard MCP channel. Once connected, you can perform document CRUD operations, manage collections and comments, and, when smart features are enabled, ask natural-language questions about your wiki, discover related documents, generate summaries, and get AI-suggested tags. Start by syncing your knowledge base if you plan to use RAG-based Q&A, then issue high-level prompts to retrieve, summarize, or analyze content. Use these practical patterns as you work with your LLMs:

How it works with an MCP client

• Use the MCP client to send requests for documents, collections, and comments just like you would with any API-backed knowledge base. - If smart features are enabled, you may issue natural-language prompts such as “Summarize the onboarding guide” or “Find documents related to deployment policies” and receive AI-generated results with source references. - Maintain authentication and URL endpoints in your client configuration to ensure secure access to your Outline instance.

Sample workflows you can run

• Query policy guidance in natural language and get an answer with links to source documents. - Discover semantically related documents to a given topic, not just keyword matches. - Generate concise summaries for long documents to speed up reviews. - Get AI-recommended tags to improve future searchability.

Available tools

search_documents

Search documents by keyword with pagination

get_document_id_from_title

Find document ID by title

list_collections

Get all collections

get_collection_structure

Get document hierarchy in a collection

list_recent_documents

Get recently modified documents

get_document

Get full document content by ID

export_document

Export document in Markdown

create_document

Create a new document

update_document

Update document (supports append)

move_document

Move document to another location

archive_document

Archive a document

unarchive_document

Restore archived document

delete_document

Delete document (soft/permanent)

restore_document

Restore from trash

list_archived_documents

List archived documents

list_trash

List trashed documents

add_comment

Add comment (supports replies)

list_document_comments

Get document comments

get_comment

Get specific comment

get_document_backlinks

Find linking documents

create_collection

Create collection

update_collection

Update collection

delete_collection

Delete collection

export_collection

Export collection

export_all_collections

Export all collections

batch_create_documents

Create multiple documents

batch_update_documents

Update multiple documents

batch_move_documents

Move multiple documents

batch_archive_documents

Archive multiple documents

batch_delete_documents

Delete multiple documents

smart_status

Check status and indexed count

sync_knowledge

Sync docs to vector database

ask_wiki

RAG-based Q&A on wiki content

summarize_document

Generate AI summary

suggest_tags

AI-suggested tags

find_related

Find semantically related docs

generate_diagram

Generate Mermaid diagrams