leepokai/jev-guard
67 9JavaScriptMITUpdated 3 weeks ago
Auto mode for every coding agent, built on Jev: risk-scores every tool call with session context (deny / ask / allow), flags prompt injection in results, checks skills and plugins. Claude Code, Codex, Copilot, Gemini, Cursor, pi, OpenCode, ACP.
ai-agentsclaude-codecodexcursorgemini-clihooksjevopencodeprompt-injectionsecurity
Our verdict: worth using
This repository provides a security hook for coding agents that evaluates tool calls and flags potential risks, which is highly relevant for enhancing the safety of agent interactions.
Filed under security in our directory.