Skip to content

Claude Skills by safetycard

github.com/safetycard

1 skillB× 10 installs0 views

SafetycardBSafetyCard: produce a defense-in-depth score for an AI agent, agent framework, coding agent, or MCP/tool server from its source code: 0.0-10.0 across 10 OWASP-Agentic-aligned criteria (1.0 each, graded on strength, coverage, defaults, and blast radius) that credit the safeguards built into the agent. Every rating is anchored to verifiable file:line evidence at a pinned commit and rendered as a markdown report and a website-ready JSON entry. Use this skill whenever the user mentions SafetyCard...securitypythonrustVotes: 0GitHub stars: 3