Agent-Blackbox: git blame for AI agents – find who broke prod in 3s
I built Agent-Blackbox after losing a week debugging an 8-agent pipeline failure. Every agent pointed fingers at the next one, and logs were useless for tracing the actual decision chain.
What it does:
- 📍 Traces failures to the exact agent and decision point
- 🔗 Reconstructs the full task-based-on chain across agents
- ✅ Uses cryptographically signed receipts (JEP / JAC) so evidence isn't just "trust me bro"
How it works:
Four verbs—`J` (Judge), `D` (Delegate), `T` (Terminate), `V` (Verify)—model any accountability workflow. Every action produces a signed receipt. When something breaks, you follow the `task_based_on` chain back to the failure.
Stack:
- Rust core + Python SDK (TypeScript coming)
- IETF drafts: JEP & JAC
- MIT licensed, fully open source
If you've ever wasted hours grep'ing through agent logs or playing detective across a multi-agent pipeline, I'd love to hear how you currently handle it. Also open to feedback on the JEP/JAC approach—does cryptographic accountability matter in your stack?
Replies