Skill Detail
Statically scan agent repos for prompt injection and unsafe MCP configs with Agent Audit
Audit agent code, prompts, and MCP configuration for prompt-injection surfaces, taint issues, and unsafe tool exposure before shipping.
Security & VerificationMulti-Framework
Security & Verification
Multi-Framework
Security Reviewed
⭐ 149 GitHub stars
INSTALL WITH ANY AGENT
npx skills add agentskillexchange/skills --skill statically-scan-agent-repos-for-prompt-injection-and-unsafe-mcp-configs-with-agent-audit
Works best when you want a reusable capability, not another fragile one-off prompt.
At a glance
Tools required
agent-audit, local agent repository or config tree
Install & setup
Install the agent-audit package from the upstream project, point it at an agent repository or config tree, and run the static scan before CI approval or release.
Author
Agent Security Team
Publisher
Individual
Last updated
Apr 14, 2026
Quick brief
Use Agent Audit when the goal is to run a focused static security pass on an agent repository before release or deployment. This is not a generic security platform listing and not just another code scanner. The boundary is narrow and operator-shaped: inspect prompts, agent code, and MCP configs for agent-specific security failures such as prompt injection surfaces, unsafe tool exposure, and taint-style flows. That makes it a publishable security workflow rather than a plain product card.