agrag.agents.verification.verify_findings

verify_findings(model:Any, question:str, sub_questions:Sequence[str], findings:str) -> VerificationResult

Run the verifier on fixed inputs and return its structured verdict.

Runs the fixed-input verifier used by the calibration evaluator. The default task agent does not use this helper or a verifier subagent.

Parameters:

• model (Any) – A LangChain chat model.
• question (str) – The original question.
• sub_questions (Sequence[str]) – The sub-questions the question was split into.
• findings (str) – The findings with citation keys such as E1, and the evidence text for each key.

Returns:

• VerificationResult – The verifier's verdict.

Raises:

• ValueError – The model returned no verdict, after the agent asked again a few times.