grith
Runs underneath AI coding agents on Linux, checks risky system calls before they run, and keeps an audit log of each decision.
Details
- What it does
- Security and passwords, Developer tools
- Who it’s for
- Developers, Software companies
- Pricing
- Free plan
- Works on
- Linux
- Extras
- Open source
- Works with
- claude code, codex, aider, cline, goose
Similar software
- Clampd
A self-hosted proxy that checks every tool call an AI agent makes against policy and blocks risky ones.
Since 2026
Free plan · Security and passwords · Developers
- DashClaw
Holds risky actions from unattended AI coding agents until someone approves them from a dashboard or phone, and logs each decision.
Since 2026
Free trial · Security and passwords · Developers
- Lelu
Checks each action an AI agent tries against your rules, asks a human for risky ones, and logs every decision.
Since 2026
Security and passwords · Developers
- SigmaShake
Checks each tool call from AI coding agents against your rules and blocks or rewrites risky ones before they run.
Since 2026
Subscription · Security and passwords · Developers