ctx_feedback Workflow & Orchestration Harness feedback for LLM output tokens/latency (local-first). Actions: record|report|json|reset|status.
Parameters
9
Required
0
Actions
5
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
action | "record" | "report" | "json" | "reset" | "status" | No | Operation to perform (default: report) |
agent_id | string | No | Agent ID (optional; defaults to current agent when available) |
intent | string | No | Intent/task string (optional) |
latency_ms | integer | No | Optional for action=record |
limit | integer | No | For report/json: how many recent events to consider (default: 500) |
llm_input_tokens | integer | No | Required for action=record |
llm_output_tokens | integer | No | Required for action=record |
model | string | No | Model identifier (optional) |
note | string | No | Optional note (no prompts/PII) |
Actions
record ctx_feedback action="record"
report ctx_feedback action="report"
json ctx_feedback action="json"
reset ctx_feedback action="reset"
status ctx_feedback action="status"
Schema MD5:
c8f0abaf207d048062889c97b9214537 ← Back to all tools