What is your agent really doing?

Atested opens the black box. Every action your AI takes must have a reason to move forward. Every ALLOW and every DENY is recorded.

Hallucinated a reason to delete that data file? Decided you just have to have a new Mac Mini? Not on Atested's watch. Have an edge case that genuinely needs to run? You approve it, once and done.

Running OpenClaw, Claude Code, or any AI agent? Point it through Atested. Every model-proposed action is classified and recorded before it reaches your system.

Install from GitHub How it works

Transparent

See everything

Every action your AI agent proposes is visible before it executes. Every decision is recorded with the evidence that drove it. Nothing happens in the dark.

Deterministic

Replace judgment with method

Where deterministic methods exist, Atested uses them. Where judgment remains, Atested is honest about it.

Verifiable

Prove it happened

Ed25519 signed, hash-chained records on disk. Anyone with the public key can verify the entire history on their own machine. Mathematically tamper-evident.

Invisible

Zero token use

Atested sits on the connection, not in the conversation. Zero token use, imperceptible latency. The model has no awareness of Atested. Denied actions return standard responses the model handles normally. Full capability preserved.

All traffic passes through Atested