Jev in production › Agent guardrails and approvals

guardrails-md

Opencode and Claude Code plugin scores every bash command with Jev in about 100ms, blocking destructive or credential-leaking calls (guardrails-md).

Open on GitHub ↗

100msmeasured, as published by the source
Use
Agent guardrails and approvals
Industry
Security
Form
Open-source tool
Stage
In production
Plugs into
Opencode, Claude Code
Listed
2026-10-22
Found via
github
Repository
berget-ai/guardrails-md
Stars
11
Forks
2
Last push
2026-10-22
Language
TypeScript
License
MIT

The README opens with

Stops your coding agent from running the bash command you'd regret.

Every bash command your coding agent (opencode, pi or Claude Code) is about to execute is scored by a small decision model first, in about 100 ms. If the command destroys data, leaks a secret, or breaks a rule in your repo's GUARDRAILS.md, the call is blocked and the agent is told why, so it can pick another route.

Teams that work close to production. The agent runs kubectl, helm and git push all day, and the difference between routine and incident is your team's policy, not a filename pattern. The gate reads that policy from GUARDRAILS.md and applies it to every command.

Badge

For the project's own README, linking back here:

Listed in Jev in production

Also used for agent guardrails and approvals