Check if the local OpenEvidence session is valid.

List question history from OpenEvidence account.

Fetch article payload by article id.

Create a question and optionally wait for completion. For follow-up question pass original_article_id.

List all collections owned by the authenticated user.

Fetch a collection (incl. nested questions[] = membership list) by id.

Create a new collection. By convention, agent-managed names start with '#'.

Add a chat (article) to a collection. Idempotent in practice.

Create the local SQLite mirror at $OE_MCP_DB_PATH (default ~/.openevidence-mcp/db/oe.sqlite). Idempotent.

Paginate /api/article/list and upsert chats. Incremental by default (stops on the first all-known page).

Refresh collections and memberships from the API into local SQLite. Prunes collections + memberships the server no longer reports.

Chats with no membership in any '#'-prefixed collection. Returns {unsorted_count, shown, items[]}.

Counts (chats, collections, hashtag, memberships, unsorted) + last sync timestamps.

Predict hashtags for every unsorted chat (or all chats with reclassify_all). Combines curated keyword rules with a per-tag log-odds-ratio signature trained from your existing memberships. Returns the proposed plan as JSON; use oe_collections_bulk_apply to actually write.

For each {article_id, hashtags[]} entry: ensure each '#'-prefixed collection exists, then add the article to it. Idempotent. Refuses non-'#' tags.