Workfile MCP server

Workfile is a helper that keeps your project's to-do cards, documents, changelog and team notes as simple Markdown files inside your code repository. It is made for teams where people and AI assistants both work on the same project and need one shared place to write things down. If you already use ChatGPT or Claude for planning and writing, this connects that work to files your team can actually read and review.

What this MCP server does

You ask your AI something like "show me card T-0042" or "add a learning about our release process". The AI passes that request to the Workfile helper running on your computer. The helper reads or writes the matching Markdown file in your repository's .project folder. Then it sends the answer back, and your AI shows it to you in the chat. Because everything is stored as files, your team can review changes in a normal pull request.

Click to zoom

What you can do with it

• List and search project cards, docs, changelog entries and memory notes
• Create a new card, document or learning note with a title and area
• Claim a card so two AI helpers do not edit the same one at once
• Move a card through statuses like backlog, doing, review and done
• Add a changelog fragment for an upcoming release
• Look up decisions, incidents and conventions recorded by the team
• Check the workspace for broken links or missing information

Try asking your AI

• “List all cards in the infra area that are still in progress”
• “Create a card titled Update billing docs in the docs area”
• “Show me the last three decisions we recorded about our database”
• “Add a changelog entry for the new search feature”

What it gives back to you

You get back plain answers in the chat: lists of cards, the full text of a document, search results, or a short confirmation that a file was created or updated. When your AI changes something, it tells you which record it touched, like card T-0042 or decision ADR-0009. The actual change lives in a Markdown file in your repository, so you can open it and read it yourself.

Before you start

Who it's for

Product managers, team leads and developers who plan work in a code repository and want their AI assistant to read and write those same project files.