2.6 KiB
2.6 KiB
Evidence: chat-verifier-agent
Code Evidence
Complex chat path now invokes verifier deterministically
- File:
src/main/java/com/superbiz/agent/service/ChatService.java - Evidence:
executeChatComplexcalls planner, executor, then verifier directly throughcallAgent(...). - Conclusion: runtime no longer depends on prompt-only Supervisor behavior to call verifier.
Verifier receives explicit inputs
- File:
src/main/java/com/superbiz/agent/hook/VerifierInputHook.java - Evidence: the hook builds a JSON payload with
original_query,executor_final_answer,tool_trace_summary, andretry_context. - Conclusion: verifier input is stable and does not depend on guessing the last assistant message from raw history.
Tool evidence is traceable to persisted invocations
- File:
src/main/java/com/superbiz/agent/service/ToolTraceSummaryService.java - Evidence: summaries include
trace_ref,source_invocation_ids,query_samples,retrieval_layers,relevance_levels, andsource_documents. - Conclusion: verifier facts can be correlated with actual
tool_invocationrows.
Verifier facts preserve evidence references
- File:
src/main/java/com/superbiz/agent/service/ChatService.java - Evidence: verifier parsing preserves
facts_checked[*].evidence_refsand persiststool_trace_summaryunderverifier_evaluation. - Conclusion:
self_evaluationnow contains both verifier judgments and the evidence index used to form them.
Evaluation channels no longer overwrite each other
- File:
src/main/java/com/superbiz/agent/service/SelfEvaluationMergeService.java - Evidence: rule and verifier evaluations are merged into separate keys.
- Conclusion: asynchronous rule scoring preserves verifier output.
Verifier logging is less noisy
- File:
src/main/java/com/superbiz/agent/hook/AgentLoggingHook.java - Evidence: verifier
thoughtstores a concise verdict summary, while fuller model output remains available in structured storage. - Conclusion:
agent_step.thoughtis no longer a misleading place for full verifier JSON.
Runtime Evidence
- Compile verification passed:
mvn -q -DskipTests compile. - Runtime session
9138f064executedplanner -> executor -> verifier. - Runtime session
9138f064persistedverifier_evaluation.facts_checked[*].evidence_refs. - Runtime session
9138f064persistedverifier_evaluation.tool_trace_summary[*].source_invocation_ids.
Design Evidence
LOW_CONFIDreturns a fixed disclaimer and verifier-derived gaps.REJECTreturns degraded output and does not pass through the raw Executor answer.retry_contextis derived from verifier-identified missing evidence facts.