Add Metric Failure Mode Insight Category Example Calls to Labs

Sample the latest calls in one persistent failure category and add them to Labs for its metric, with feedback as the note and an optional corrected expected_value label (defaults to a pass).

POST /observability/v1/metric-failure-mode-insights/categories/{category_slug}/fix-metric/

Add an accumulating failure category's example calls to Labs

cURL

curl --request POST \
  --url https://api.cekura.ai/observability/v1/metric-failure-mode-insights/categories/{category_slug}/fix-metric/ \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "failure_mode_index": 1,
  "feedback": "<string>",
  "expected_value": "<unknown>"
}
'

200

{
  "failure_mode_index": 1,
  "feedback": "<string>",
  "expected_value": "<unknown>"
}

Authorizations

oauth2supabase_sessionembedded_sessionapi_key

Authorization · string · header · required

OAuth access token issued by Cekura for connected apps.

Path Parameters

category_slug · string · required

Pattern: ^[^/]+$

Body

application/jsonapplication/x-www-form-urlencodedmultipart/form-data

Request body for POST /metric-failure-mode-insights/{id}/fix_metric/.

Samples a few of one failure mode's example calls and adds them to Labs for the insight's metric, carrying the user's feedback as the note and expected_value as the corrected label.

failure_mode_index · integer · required

Index into the insight's failure_modes list to act on.

Required range: x >= 0

feedback · string · required

Human note attached to each sampled call's metric review in Labs.

expected_value · any | null

Corrected label to record for each sampled call in Labs: true/false for binary metrics, a 0-5 score for ratings, a number for numeric metrics, or null for N/A. Defaults to a passing value when omitted.

Response

200 - application/json

Request body for POST /metric-failure-mode-insights/{id}/fix_metric/.

Samples a few of one failure mode's example calls and adds them to Labs for the insight's metric, carrying the user's feedback as the note and expected_value as the corrected label.

failure_mode_index · integer · required

Index into the insight's failure_modes list to act on.

Required range: x >= 0

feedback · string · required

Human note attached to each sampled call's metric review in Labs.

expected_value · any | null

Corrected label to record for each sampled call in Labs: true/false for binary metrics, a 0-5 score for ratings, a number for numeric metrics, or null for N/A. Defaults to a passing value when omitted.