Claude Skill MartBrowse skillsQuick linesLearn by videoTerminal guideWhat is a Skill?
 
← Back to list

Provenance Evidence Chain (Taishige)

Collects every operation event across a multi-agent system and turns it into a trace_id-linked evidence chain plus vector index for audit and replay.

AutomationAdvanced★ 6,083⑂ 585AI score 5/10Last updated: Sep 2, 2026

What it does

  • Captures agent actions — dispatch, scan, parse, card generation — as structured events.
  • Normalizes them into actor / action / target / timestamp / trace_id and builds a traceable evidence_chain.
  • Writes to Qdrant (vector index via MCP) with SQLite as a local fallback, enabling semantic search plus fault tolerance.
  • Exposes a query API by trace_id, time window, or agent, powering audit dashboards and rollback localization.
  • Defines explicit failure behavior: vector write failure → cache in SQLite and backfill asynchronously; invalid event schema → drop that record only; query timeout → return the last good snapshot marked stale.

Who it's for

  • Engineers designing execution logging and audit trails for multi-agent pipelines.
  • Teams that need to replay agent decisions for debugging or compliance.
  • Infra owners already running (or considering) a Qdrant + SQLite observability stack.
  • Note: it assumes the repo's own architecture (core.runtime.AgentSession, memory.taishige), so you need that codebase to run it as-is.

Examples

  1. Run python skills/provenance/scripts/run_provenance.py → produces trace.jsonl and trace_summary.json under examples/snse_survey/provenance/.
  2. Query one trace_id to see which stage corrupted an artifact and pick a rollback point.
  3. Simulate a Qdrant outage to verify events are cached in SQLite and backfilled after recovery.

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/anbeime/skill/HEAD/antinet-agentteams/skills/provenance/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the antinet-agentteams/skills/provenance folder from the GitHub repo anbeime/skill into my ~/.claude/skills/provenance/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/anbeime/skill.git && mkdir -p ~/.claude/skills && cp -r skill/antinet-agentteams/skills/provenance ~/.claude/skills/provenance

⚠ This is a third-party skill. Check the source repository before installing.

  1. Open a terminal and move to your working directory.
  2. Clone the repo: git clone https://github.com/anbeime/skill.git
  3. Create the skills folder: mkdir -p ~/.claude/skills
  4. Copy the skill: cp -r skill/antinet-agentteams/skills/provenance ~/.claude/skills/provenance
  5. To use the real entry point, go to the repo root (skill/antinet-agentteams) and run python skills/provenance/scripts/run_provenance.py to confirm outputs are generated.
  6. For vector search, prepare Qdrant (via MCP), Python 3, and SQLite beforehand.
  7. Restart Claude Code and ask something like "use the provenance skill to log this run's event chain".
View source on GitHub ↗