Agents & Machines

`gradeRun`

Grade Run

Grade one of your agent runs: record whether it was "useful", "partial", or "failed", plus an optional failure mode when it fell short ("grounding" | "instruction-following" | "format" | "tool-use"…

write

Arguments

failureModestringoptionalThe failure mode when the run fell short: the cultivation target (which harness weakness to fix). "grounding" (unsupported claims), "instruction-following" (ignored the ask), "format" (wrong shape/structure), "tool-use" (never called a…
outcomestringrequiredThe task outcome: "useful" (did the job), "partial" (partly there), or "failed" (missed the task: still worth grading for the failure mode).
sessionIdstringrequiredThe run to grade: the conversationId of the session (from listAgentSessions or getAgentResult; sessionId IS the conversationId).

Returns

The registry does not declare a return shape.

Ask for it

The Cloud routes these requests to this verb.

grade this run
mark that run useful/failed
this run was a partial
the agent failed on tool-use

Pair it with

The verbs that finish the job.