provenance

Provenance Tracing Skill (Taishige)

Usage

• Called by each government office at key operation nodes (or pushed asynchronously via the event bus), writing into the evidence chain.
• Provides retrieval interfaces by trace_id / time window / government office dimension, supporting the audit panel and rollback localization.

Input

• event: operation event, including actor(government office name),action、target、timestamp、trace_id
• payload: (optional) structured artifact reference associated with the event

Output

• evidence_chain: traceable evidence chain linked by trace_id
• vector_index: vector index entries for semantic retrieval
• query_api: interface description for retrieving historical evidence by condition

Dependencies

• Qdrant (vector storage, accessed via MCP)
• SQLite (structured event persistence, local fallback)
• Event reporting protocol of each government office (unified schema)

Failure Handling

• Vector database write failure → cache the event locally in SQLite, asynchronously backfill after recovery, without blocking the main flow.
• Single event schema invalid → record and discard that entry, without affecting the writing of the whole chain.
• Retrieval timeout → return the most recent successful snapshot and annotate stale。

Reuse Value

• Observability foundation: any multi-Agent system can directly mount it and gain out-of-the-box auditing and replay capabilities.
• Review fit: a natural scoring point for the "Engineering Implementation and Operational Validation and Security Auditing (20%)" dimension of the Agent Infra track.

Runnable package execution (runnable package)

• Real entry point:scripts/run_provenance.py
• Execution is equivalent to core.runtime.AgentSession.run_stage("provenance"), calling memory.taishige.TaiShiGeAgent.writeback(flowing back the full-link events and four-color cards into the evidence chain, pure Python and offline-capable).
• Run:python skills/provenance/scripts/run_provenance.py
• Artifacts:examples/snse_survey/provenance/(trace.jsonl + trace_summary.json, containing every Junjichu dispatch and government office execution event).