Skip to content

v2.0.0 — AI Security Tools: scan_skill & check_prompt

Choose a tag to compare

@alberthild alberthild released this 03 Mar 17:12
· 16 commits to main since this release

🆕 New MCP Tools

check_prompt — Prompt Injection Detection

Detect prompt injection in text before processing untrusted input. 208 patterns, multi-language, 4 decoders. <100ms response time.

scan_skill — Skill Security Scanner

Scan AI agent skills for supply chain attacks. 204 patterns, 8 risk categories (Snyk ToxicSkills taxonomy). Automatic secret redaction.

All Tools

check_prompt, scan_skill, check_url, check_password, check_domain, check_ip, check_email, full_scan

x402scan Discovery

ShieldAPI is now registered on x402scan.com.

Install

npm install -g shieldapi-mcp

Links