mentor operations

The exact fields each mentor.* operation takes and returns.

Every op is POST /v1/<op> over HTTP and the MCP tool loom_<op>. The identity fields (org, namespace, agent, user_id, session_id) are filled from the token and the MCP session when a transport can; for an HTTP call with no token on a single-user server, the install's org on every op but code.* and tracker.* (which keep their plane's), and on some ops its namespace and agent; see identities and namespaces. The operations index explains role, kind and the MCP all listing. Unknown input fields are served and named in ignored_args.

mentor.brief

One agent's usage profile and the store's hints.

POST /v1/mentor.brief · MCP loom_mentor_brief · role viewer · read

Input field	Type	Default	Meaning
org	str	required	The tenant (first element of the identity tuple).
namespace	str	required	The namespace the op reads or writes.
agent	str	''	The writing or reading agent; '-' or '*' means any agent where the op allows it.
window_seconds	int	604800	How far back the ledger read reaches, in seconds; 1 hour to 90 days.

Output field	Type	Meaning
org	str	The tenant (first element of the identity tuple).
namespace	str	The namespace the op reads or writes.
agent	str	The writing or reading agent; '-' or '*' means any agent where the op allows it.
window_seconds	int	The window the profile was computed over, in seconds; a profile served from the cache keeps the window it was computed with.
computed_at	datetime	When the profile was computed (UTC); older than the call when served from the cache.
sessions	int	Distinct client-named sessions of this agent in the window (- and * not counted).
usage	dict[str, int]	Calls per operation by this agent in the window.
planes	list[MentorPlane]	Every plane the mentor knows, with this agent's calls to it and whether it is available.
hints	list[MentorHint]	The rules that fired for this agent, each with the counts that fired it; empty when none.
adoption	list[MentorAdoption]	Per hint code, sessions it was shown in and followed in, most shown first; empty when no hint reached the agent or the read failed.
usage_evidence	list[UsageRung]	The agent's use of Loom in the window, one rung each, weakest claim first: from a call made to a graded run.
ledger_from	datetime \| None	Oldest ledger row of this agent in the window, of any kind; None when it holds none or the read failed.
ledger_retention_days	int \| None	Days the configured retention keeps this namespace's call ledger rows; 0 keeps them until deleted, None when the read failed.
agents	list[str]	The namespace's writers in the window, most active first; the brief defaults to the first when the caller has no agent name of its own (the dashboard) — agent_defaulted says so.
agent_defaulted	bool	True when the caller named no real agent and the brief picked the top writer in agents.
skipped	list[str]	Facts the brief could not read, as "step: reason".
healthy	bool	No warning or critical hint.

mentor.guide

The agent guide this server ships; the same for every caller.

POST /v1/mentor.guide · MCP loom_mentor_guide · role viewer · read · not listed under the MCP all toolset

Input field	Type	Default	Meaning
org	str	''	The tenant (first element of the identity tuple).
part	Literal['skill', 'reference', 'all']	'skill'	Which part: skill (the loop and rules, the default), reference (each plane) or all.

Output field	Type	Meaning
version	str	The guide's version (metadata.version of skills/loom/SKILL.md).
part	Literal['skill', 'reference', 'all']	The part returned, as requested.
text	str	The guide as Markdown, without the skill's frontmatter.