antinet-provenance

Collects multi-agent operation events into traceable evidence chains with vector indexes.

6.0k|578|Updated Feb 2, 2026
One-click install
npx skills add https://github.com/anbeime/skill --skill antinet-provenance
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: antinet-provenance
Source: https://github.com/anbeime/skill/tree/main/skills/antinet-provenance
Command: npx skills add https://github.com/anbeime/skill --skill antinet-provenance

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Multi-agent systems often lack end-to-end observability, making it impossible to audit who did what, when, and why. This Skill records every dispatch, scan, parse, and card-generation event into a traceable evidence chain so agent runs become auditable and replayable.

Core Features & Use Cases

  • Evidence Chain Generation: Links events by trace_id into a queryable provenance chain stored as trace.jsonl and trace_summary.json.
  • Vector Indexing: Writes semantic index entries to Qdrant for retrieval, with SQLite as a local fallback when the vector store fails.
  • Resilient Failure Handling: Caches events locally on vector-store failure, drops malformed events without breaking the chain, and returns stale snapshots on query timeout.
  • Use Case: Run the provenance stage after a multi-agent pipeline to produce a full audit trail of every agent dispatch and execution event for security review or rollback analysis.

Quick Start

Run the provenance skill with python scripts/run_provenance.py to execute the full agent pipeline and write the trace evidence chain to the examples output directory.

Frequently Asked Questions about antinet-provenance

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I add audit logging to a multi-agent system?

Emit structured events with actor, action, target, timestamp, and trace_id at each key operation node, then write them into an evidence chain store. This Skill chains events by trace_id and produces trace.jsonl plus a summary for audit panels.

How do I run the provenance skill script?

Run python scripts/run_provenance.py from the skill directory. It triggers the full agent pipeline stages and writes the evidence chain output to examples/snse_survey/provenance/ as trace.jsonl and trace_summary.json.

What happens when the Qdrant vector store write fails?

Events are cached in a local SQLite database and asynchronously backfilled after recovery, so the main pipeline is never blocked. Malformed individual events are logged and discarded without breaking the rest of the chain.

Can I query provenance events by trace ID or time window?

Yes, the skill exposes a query interface supporting retrieval by trace_id, time window, or actor dimension. On query timeout it returns the most recent successful snapshot marked as stale.

What are the limitations of event-sourced agent tracing?

It depends on all agents emitting events under a unified schema, so non-compliant producers create gaps in the chain. It also adds storage overhead for vector indexes and event logs in high-throughput systems.