Mnestiq open-sources AI agent forensic logging tool
Understand this story
Mnestiq has released and open-sourced AI agent forensic logging tool v1.1.1, which can trace connections in firewall or traffic logs back to a specific agent's run and step, and records the inputs, decisions, and approvals the model saw. Records are sealed with a hash chain and Ed25519-signed Merkle checkpoints, so tampering, deletion, or reordering will cause verification to fail. The tool is licensed under Apache-2.0, installable via pip, requires Python 3.10+, supports Anthropic, OpenAI, Azure OpenAI, and compatible APIs, and includes an evidence format, logger, verifier, and local dashboard. Mnestiq says prompts and data stay in the user's own systems, and only signed checkpoint fingerprints are sent to the user-configured destination. Mnestiq says it is looking for a small number of security and platform teams to refine the workbench around real incidents.
AI overview based on source reports · Updated 19 min ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Hacker News AIShow HN: Link firewall log entries to the AI agent step behind them
Mnestiq is an open-source (Apache-2.0) Python recorder that captures what AI agents saw, decided and did as hash-chained records sealed by Ed25519-signed Merkle checkpoints, so a firewall or flow-log connection can be traced to the specific agent run, step, input and approval behind it. It supports Anthropic, OpenAI, Azure OpenAI and compatible APIs, and ships with a verifier and local dashboard. The project states its limits plainly: it proves records weren't altered after writing, not that a compromised host told the truth.
Activity for this story
Not enough continuous observations to show a trend yet.