list_knowledge_entries
Page what a project has taught Subscriby Copilot — every Knowledge entry with its kind, audience and state, never its text.
Read what a project has written down for Subscriby Copilot: the questions with their answers and the documents, newest first, each with its kind, who may be answered from it (audience), whether its passages are in the index (status), its title and how many passages it became. The rows carry no text, so an agent can read a project's whole Knowledge in one call and ask for the entries it needs with get_knowledge_entry.
audience is the member boundary. A members entry may be read to a member in the inbox, the
project bot and the chat on the portal page; a team entry serves the creator's own surfaces alone,
the suggested reply in the inbox first. Nothing else about an entry changes with the audience.
Requires ability
The token behind the MCP session must hold it, or the call is refused with TOKEN_MISSING_ABILITY.
Runs the same action as
The REST endpoint and this tool share one action, so validation, permissions and events are identical.
Annotations
It reads and never changes anything.
Arguments
project_id*stringUUID of the project whose Knowledge to list.
limitintegeroptionalMaximum rows to return per page (1..100).
1max100pageintegeroptional1-indexed page number.
1What it returns
{ "data": [ { "id": "3b7e4d2a-9c1f-4a6b-8d5e-2f7c9a1b3e4d", "project_id": "7f3d1c92-8b45-4e6a-9d21-5c8e0a4b6f13", "kind": "question", "kind_label": "Question & Answer", "audience": "members", "audience_label": "Members & Team", "status": "ready", "status_label": "Ready", "title": "When does my membership renew?", "body": null, "file_name": null, "file_size": null, "mime": null, "chunk_count": 1, "characters": 64, "failure": null, "indexed_at": "2026-10-19T16:02:11+00:00", "created_at": "2026-10-19T16:02:11+00:00", "updated_at": "2026-10-19T16:02:11+00:00" } ], "meta": { "project_id": "7f3d1c92-8b45-4e6a-9d21-5c8e0a4b6f13", "page": 1, "limit": 25, "total": 1, "has_more": false }}body is null on every row of the list. kind is question or document; status is ready once the passages are in the index, indexing while they are cut, failed when a document's text could not be read, with failure saying why. file_name, file_size and mime are set only for a document uploaded as a file on the dashboard; a document sent as text has none.
How it fails
RESOURCE_NOT_FOUNDproject_id names a project the token cannot see: unknown, another team's, or outside the token's scope:project: allow-list.
AUTHENTICATION_REQUIREDno authenticated user on the request.
TOKEN_MISSING_ABILITYtoken lacks project-knowledge:view-any.
How is this guide?
Last updated on