Config evaluate golden

Candidate config scored against the standing golden set.

POST /v1/brands/{brand_id}/config/evaluate-golden

cURL

curl --request POST \
  --url 'https://api.workstream.stateset.com/v1/brands/{brand_id}/config/evaluate-golden' \
  --header "x-api-key: $STATESET_TEMPORAL_API_KEY" \
  --header 'Content-Type: application/json' \
  --data '{
  "config": {},
  "provider": "string",
  "model": "string",
  "pass_threshold": 1.5,
  "limit": 20,
  "simulate_tools": true
}'

200

{
  "gated": true,
  "passed": true,
  "threshold": 1.5,
  "case_count": 1,
  "summary": {
    "total": 102,
    "passed": 1,
    "pass_rate": 7.5,
    "avg_score": 7.5,
    "by_dimension": {},
    "flags": {},
    "regressions": [
      {
        "case_id": null,
        "baseline_score": null,
        "candidate_score": null,
        "delta": null
      }
    ],
    "action_fidelity": {}
  },
  "cases": [
    {
      "case_id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
      "draft": "string",
      "tool_calls": [],
      "action_fidelity": "string",
      "verdict": {
        "score": null,
        "passed": null,
        "dimensions": null,
        "rationale": null,
        "flags": null
      }
    }
  ],
  "reason": "Two-person tent, green — replacement for damaged pole set."
}

Path parameters

brand_id · string (uuid) · required

Request body

object

config · object

provider · string

model · string

pass_threshold · number

limit · integer

simulate_tools · boolean

Response

object

gated · boolean · required

passed · boolean · required

threshold · number · required

case_count · integer · required

summary · object

Show summary

total · integer · required

passed · integer · required

pass_rate · number · required

avg_score · number · required

by_dimension · object · required

flags · object · required

regressions · object[] · required

Show regressions

case_id · string · required

baseline_score · number · required

candidate_score · number · required

delta · number · required

action_fidelity · object

Present only when non-empty

cases · object[] · required

Show cases

case_id · string · required

draft · string · required

tool_calls · string[] · required

action_fidelity · string

verdict · object · required

Show verdict

score · number · required

passed · boolean · required

dimensions · object · required

rationale · string · required

flags · string[] · required

reason · string

Status codes

Code	Meaning
200	Candidate config scored against the standing golden set.
400	Validation error
401	Unauthorized
404	Not found
429	Rate limited
500	Internal error

cURL

curl --request POST \
  --url 'https://api.workstream.stateset.com/v1/brands/{brand_id}/config/evaluate-golden' \
  --header "x-api-key: $STATESET_TEMPORAL_API_KEY" \
  --header 'Content-Type: application/json' \
  --data '{
  "config": {},
  "provider": "string",
  "model": "string",
  "pass_threshold": 1.5,
  "limit": 20,
  "simulate_tools": true
}'

200

{
  "gated": true,
  "passed": true,
  "threshold": 1.5,
  "case_count": 1,
  "summary": {
    "total": 102,
    "passed": 1,
    "pass_rate": 7.5,
    "avg_score": 7.5,
    "by_dimension": {},
    "flags": {},
    "regressions": [
      {
        "case_id": null,
        "baseline_score": null,
        "candidate_score": null,
        "delta": null
      }
    ],
    "action_fidelity": {}
  },
  "cases": [
    {
      "case_id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
      "draft": "string",
      "tool_calls": [],
      "action_fidelity": "string",
      "verdict": {
        "score": null,
        "passed": null,
        "dimensions": null,
        "rationale": null,
        "flags": null
      }
    }
  ],
  "reason": "Two-person tent, green — replacement for damaged pole set."
}