Integration · Documents

Connect Claude & ChatGPT to Outline

Ask your AI to find the page that explains a process, sum up a document, or write a new one in your Outline wiki. It can search and read documents, look at edit history, create and update pages, add comments and move pages between collections. Your team's knowledge becomes something you can ask questions about.

Connect Outline free

Free plan, no credit card. Takes about a minute.

Try asking

“Where do we describe our onboarding process?”
The AI runs a full-text search across the wiki and shows the matching documents with text snippets.
“Summarize the Security policy page.”
It finds the document by title, reads its full Markdown text and gives you a short summary.
“Write a draft page in Engineering called Release checklist.”
It creates the document in the Engineering collection with Markdown text. It stays a draft until you publish it.

What your AI can do

Search and read

Full-text and title search, full document text, child pages, collection trees, drafts and recently viewed documents.

Write and organize

Create documents from scratch or from templates, update, publish, move, duplicate, archive and restore pages, and manage collections.

Comments and history

Read, add, edit and resolve comments, and look at a document's revisions or roll it back to an earlier one.

“What changed in the pricing page this week?”
It lists the document's revisions and reads the text of earlier versions so it can compare them.
“Reply to the open comments on the Q3 plan.”
It lists the comments on the document, adds replies in the threads and can mark a thread as resolved.
“Show me the page tree of the Handbook collection.”
It gets the collection's full structure with all nested pages.

29 tools for Outline

These tools are switched on when you connect. You can switch any of them off, or require your approval before it runs.

archive_documentVisible to othersArchive a document by `id` so it is hidden from the main navigation. It can be brought back with restore_document.
create_collectionChanges dataCreate a new collection with `name`, optional `description`, `icon`, `color` and default `permission` (read or read_write) for the team.
create_commentChanges dataAdd a comment to a document by `document_id` with Markdown `text`; pass `parent_comment_id` to reply in a thread. Mentioned people may be notified.
create_documentChanges dataCreate a document with `title` and Markdown `text` in a `collection_id` (or under a `parent_document_id`). Set `publish` true to make it visible to the team; otherwise it stays a draft. Can start from a `template_id`.
delete_documentDeletesMove a document by `id` to the trash. With `permanent` true it is deleted for good and cannot be recovered.
duplicate_documentVisible to othersMake a copy of a document by `id`, optionally with a new `title`, target `collection_id` or `parent_document_id`, and `recursive` to copy child pages too.
get_collectionChanges dataGet one collection by `id`, with its name, description and permissions.
get_collection_structureChanges dataGet the full page tree (nested documents) of a collection by `id`.
get_documentChanges dataGet one document by `id` (or URL slug id), including its title and full Markdown text.
get_revisionChanges dataGet one revision by `id`, including the document's text at that point in time.
get_userChanges dataGet one workspace member by `id`, with name, email and role.
list_archived_documentsChanges dataList archived documents, optionally within one `collection_id`.
list_child_documentsChanges dataGet the nested pages under a document, by the parent document's `id`.
list_collectionsChanges dataList the collections (top-level sections of the wiki), optionally matching `query`.
list_commentsChanges dataList comments on a document (`document_id`) or across a collection (`collection_id`).
list_document_revisionsChanges dataList the saved revisions (edit history) of a document by `document_id`.
list_documentsChanges dataList documents, optionally filtered by `collection_id`, `parent_document_id` or author `user_id`, with sort and paging.
list_draftsChanges dataList your unpublished draft documents, optionally filtered by `collection_id` or `date_filter`.
list_recently_viewed_documentsChanges dataList the documents you viewed most recently.
list_templatesChanges dataList document templates, optionally matching `query` or within a `collection_id`. Use a template's id as `template_id` when creating a document.
list_usersChanges dataList workspace members, optionally matching `query` (name or email), `emails` or `filter` (active, invited, suspended).
move_documentChanges dataMove a document by `id` to another `collection_id` and/or under a new `parent_document_id`.
resolve_commentChanges dataMark a comment thread as resolved by the comment `id`.
restore_documentChanges dataRestore an archived or deleted document by `id`, or roll it back to an earlier `revision_id`.
search_document_titlesChanges dataFind documents whose title matches `query`. Faster than full-text search when you know roughly what the page is called.
search_documentsChanges dataFull-text search across the wiki's documents with `query`, optionally narrowed by `collection_id`, author `user_id` or `date_filter` (day, week, month, year). Returns matching documents with text snippets.
update_collectionChanges dataUpdate a collection's `name`, `description`, `icon`, `color` or default `permission` by `id`. Only the fields you pass change.
update_commentChanges dataEdit the `text` of a comment by `id`.
update_documentChanges dataUpdate a document's `title` or `text` by `id`. `edit_mode` chooses whether text replaces, appends or prepends; set `publish` true to publish a draft.

Set up in three steps

1
Pick the app
Create a free PipMCP account and choose Outline from the app list.
2
Paste your key
Enter your Outline address without https:// (app.getoutline.com for Outline cloud, or your own domain if you host Outline yourself) and an API key. In Outline, open Settings › API & Apps and create a new API key. It starts with ol_api_. Treat it like a password.
3
Add the link to your AI
You get a personal MCP link. Add it to Claude, ChatGPT or Cursor:
1. Click your name, then Settings › Connectors › Add custom connector.
2. Paste your link as the Remote MCP server URL.
3. Switch it on from the + menu in a chat.

Questions

What can the AI do in Outline?

It can search and read documents, collections, templates, comments, revisions and workspace members. It can create and update documents and collections, publish drafts, move, duplicate, archive and restore documents, and add, edit and resolve comments. It can also move a document to the trash if you switch that tool on.

Does the AI see my Outline API key?

No. Your Outline API key is encrypted at rest and never shown to the AI. After you save it, it is not shown again, not even to you. The AI only sees the results of the tools it calls.

Can I control what the AI is allowed to do?

Yes. You choose which tools are switched on, so you can start with search and reading only. Deleting a document can be permanent, and it can require your approval before it runs, as can archiving or any other action you pick. Every tool call is logged.

Does it work with ChatGPT?

Yes. In ChatGPT go to Settings › Apps & Connectors › Advanced and turn on Developer mode, then add your PipMCP link. Developer mode needs a paid ChatGPT plan: Plus, Pro, Business or Enterprise. The same link also works in Claude (Settings › Connectors › Add custom connector), Cursor and other MCP clients.

What does it cost?

PipMCP has a free plan with no credit card required. Paid plans bill per completed task. You also need your own Outline account.

Does it work with self-hosted Outline?

Yes. When you connect, you enter the address of your Outline workspace. Use app.getoutline.com for Outline cloud, or your own domain if you host Outline yourself. The API key comes from Settings › API & Apps in that workspace.