jev-sentinel

A guardrail for coding agents built on Jev that checks tool calls, tool outputs, and agent replies for prompt injection, missing approvals, and leaked secrets, shipped as one shared rule set across a Pi extension, a Claude Code plugin, and a Codex CLI plugin — built on the premise that a README telling an agent to `curl | sh` is just more text the model needs to be checked against.

unverified — sourced from awesomejev.com; description independently written by AionEdge from the repo’s own GitHub metadata and README, not GitHub-API relevance-checked