Get project doc outline
get_project_doc_outlineList every markdown page in the workspace with a short summary (title, heading count, char count) plus its lifecycle — status, version and agentMayBuildFrom. An INVENTORY, not a search: NOT the way to answer a question or find the page about something — that is search_project_docs. Building this list downloads and parses every page of the site (slow on a large one), and a list of titles answers nothing by itself. It is also the REVIEW BOARD for documentation that is governed by status: status: 'review' lists what is waiting on a human, status: 'generated' what a machine wrote that nobody has read. counts always describes the whole prefix, not the filtered rows, so 'four of ninety pages are approved' is one call. Statuses: generated (A machine wrote this page and no human has read it yet.) · draft (Someone is still writing it. Not ready to be read as settled.) · review (Waiting for a human to read it and decide.) · approved (A human read this version and signed off. Safe to build work from.) · locked (Frozen on purpose. Agents may read it and build from it, but may not rewrite it.) · deprecated (Superseded. Kept so its links keep working, not to be relied on.) · archived (History. Neither built from nor edited.)
counts still covers them all. One of: generated, draft, review, approved, locked, deprecated, archived.- Use it for 'what does this documentation cover', to pick a folder for a NEW page, or after
search_project_docsfound nothing.
{
"jsonrpc": "2.0",
"id": 1,
"method": "tools/call",
"params": {
"name": "get_project_doc_outline",
"arguments": {
"status": "generated"
}
}
}