Forseti · verifiable agent audit

Verifiable agent audit, anchored to public chain.

Cryptographic proof of every action your AI agents took. Without trusting us.

Read the docsGitHub — coming soon

Tamper-evident

Every event is hashed, batched into a Merkle tree, and the root committed on-chain. Re-verify against Chia from any node, no Forseti involvement required.

Compliance-ready

EU AI Act Art. 12, DORA Art. 9–10, GDPR Art. 22 — log requirements where any tampering or gap is cryptographically detectable, while you can still erase payloads under GDPR Art. 17.

Agent-native

SDK + plugins for Claude Code, OpenClaw, Hermes, and NemoClaw. Drop in one extension and every action your agents take starts landing in a verifiable trail.