Skip to main content
GET
/
confidence-scoring
/
job
/
{job_id}
Python
from meibel import MeibelClient

client = MeibelClient(api_key="your-api-key")

result = client.confidence_scoring.get_scoring_job("job_id_value")
print(result)
{
  "job_id": "<string>",
  "agent_identity_context": {
    "customer_id": "<string>",
    "project_id": "<string>",
    "agent_name": "<string>",
    "agent_version": "<string>",
    "agent_session_id": "<string>",
    "agent_turn": 123,
    "agent_workflow_name": "<string>",
    "agent_workflow_version": "<string>",
    "agent_workflow_session_id": "<string>",
    "batch_definition_id": "<string>",
    "batch_execution_id": "<string>",
    "tool_id": "<string>",
    "tool_instance_id": "<string>",
    "tool_execution_id": "<string>"
  },
  "module": "<string>",
  "status": "<string>",
  "score": 123,
  "explanation": "<string>"
}

Authorizations

​
Meibel-API-Key
string
header
required

Path Parameters

​
job_id
string
required

Unique identifier of the scoring job to retrieve.

Response

Successful Response

A confidence scoring job record with metadata and scores only.

​
job_id
string
required

Unique identifier for this scoring job.

​
agent_identity_context
AgentIdentityContext · object
required

The agent, workflow, and tool context that produced the scored output.

​
module
string
required

The scoring module used to evaluate the output. Judge-based modules (e.g. correctness, coherence, faithfulness) produce scores on a 0–10 scale. Statistical modules (e.g. observed_consistency, data_grounding) produce scores on a 0.0–1.0 scale.

​
status
string
required

Current status of the scoring job: submitted, in_progress, completed, failed, or not_run.

​
score

The computed confidence score, or null if the job has not completed. Range depends on the module: 0–10 (integer) for judge-based modules, 0.0–1.0 for statistical modules.

​
explanation
string | null

Human-readable explanation of the score.