Sign in
risk-assessment

Update Factor Rating

Set or clear the reviewer's manual override.

Setting a value moves review_state to OVERRIDDEN and stamps the reviewer; explicitly clearing it (null) reverts to the AI's suggestion and review_state back to PROPOSED — clearing is "undo the override", not "approve", so it does not stamp an approval. reviewer_note is independent of the rating and can be set on its own.

patch/risk-assessment/factor-assessments/{factor_assessment_id}

Path parameters

factor_assessment_idstring uuid required

Request body

manual_inherent_risk_labelstring uuid nullable
reviewer_notestring nullable

Response

Successful Response

factor_assessment_idstring uuid required
status'queued' | 'running' | 'complete' | 'error' required

Lifecycle of one factor's assessment within a run.

There is no deferral state: a factor either produces a rating its rationale can defend, or it errors. Missing evidence is narrated in the rationale.

progress_pctinteger nullable required
current_activitystring nullable required
review_state'proposed' | 'approved' | 'overridden' required

Where a human reviewer stands on the agent's proposed rating.

Written by the approval workflow (M5); until then every completed factor sits at PROPOSED, which is what renders the "Kobalt proposed" chip.

reviewer_namestring nullable required
reviewed_atstring date-time nullable required
citation_countinteger required
distinct_document_countinteger required
step_countinteger nullable required
duration_msinteger nullable required
errorstring nullable required
agent_run_idstring uuid nullable required

Changes

No recorded changes to this endpoint across all 1 revision of this API.