Get Run

Get detailed run results

GET /v1/runs/{run_id}

cURL

curl https://api.playgent.com/v1/runs/run_def456 \
  -H "Authorization: Bearer your-api-key"

Response

{
  "run_id": "run_def456",
  "status": "completed",
  "success": false,
  "test_case": { "id": "tc_xyz789", "name": "Refund Request Flow" },
  "agent": { "id": "agent_abc123", "name": "Customer Support Agent" },
  "turns": [
    {
      "turn_index": 0,
      "input": [{ "role": "user", "content": "I want a refund..." }],
      "output": [{ "role": "assistant", "content": "I'd be happy to help..." }],
      "expected_behavior": "Agent should ask for order details...",
      "success": true,
      "evaluation": {
        "overall_pass": true,
        "scorer_outputs": {
          "relevance": { "score": 0.92, "reasoning": "Response directly addresses refund request" },
          "faithfulness": { "score": 0.88, "reasoning": "All claims supported by context" }
        }
      },
      "metrics": {
        "latency_ms": 1240,
        "tokens": { "input": 156, "output": 89 },
        "cost_usd": 0.0034
      },
      "trace_id": "trace_ghi789"
    }
  ],
  "summary": {
    "turns_passed": 2,
    "turns_failed": 1,
    "avg_latency_ms": 1450,
    "total_cost_usd": 0.0102
  }
}

Retrieve detailed results for a test run including all turns, evaluations, and metrics.

run_id · string · required

Unique run identifier

run_id · string · required

Run identifier

status · string · required

Run status: queued, running, completed, failed

success · boolean · required

Whether all turns passed

test_case · object · required

Test case information

Show properties

id · string

Test case ID

name · string

Test case name

agent · object · required

Agent information

Show properties

id · string

Agent ID

name · string

Agent name

turns · array · required

Array of turn results

Show turn object

turn_index · integer

Turn position

input · array

Input messages

output · array

Output messages

expected_behavior · string

Expected behavior

success · boolean

Whether turn passed

evaluation · object

Evaluation results

Show properties

overall_pass · boolean

Overall pass/fail

scorer_outputs · object

Per-scorer results with score and reasoning

grader_output · string

Grader explanation

metrics · object

Performance metrics

Show properties

latency_ms · integer

Response latency

tokens · object

Token usage

cost_usd · number

Estimated cost

trace_id · string

Linked trace ID

summary · object · required

Run summary

Show properties

turns_passed · integer

Passing turns

turns_failed · integer

Failed turns

avg_latency_ms · integer

Average latency

total_cost_usd · number

Total cost

cURL

curl https://api.playgent.com/v1/runs/run_def456 \
  -H "Authorization: Bearer your-api-key"

Response

{
  "run_id": "run_def456",
  "status": "completed",
  "success": false,
  "test_case": { "id": "tc_xyz789", "name": "Refund Request Flow" },
  "agent": { "id": "agent_abc123", "name": "Customer Support Agent" },
  "turns": [
    {
      "turn_index": 0,
      "input": [{ "role": "user", "content": "I want a refund..." }],
      "output": [{ "role": "assistant", "content": "I'd be happy to help..." }],
      "expected_behavior": "Agent should ask for order details...",
      "success": true,
      "evaluation": {
        "overall_pass": true,
        "scorer_outputs": {
          "relevance": { "score": 0.92, "reasoning": "Response directly addresses refund request" },
          "faithfulness": { "score": 0.88, "reasoning": "All claims supported by context" }
        }
      },
      "metrics": {
        "latency_ms": 1240,
        "tokens": { "input": 156, "output": 89 },
        "cost_usd": 0.0034
      },
      "trace_id": "trace_ghi789"
    }
  ],
  "summary": {
    "turns_passed": 2,
    "turns_failed": 1,
    "avg_latency_ms": 1450,
    "total_cost_usd": 0.0102
  }
}