Aguara — Open-Source Security Scanner for AI Agent Skills & MCP Servers

# Aguara > Open-source static security scanner for AI agent skills and MCP servers. 148+ detection rules across 13 categories. Scans skill files, tool schemas, and server configurations for prompt injection, credential exfiltration, supply chain attacks, and more. 100% local, zero dependencies, SARIF output. ## Core Capabilities - [Security Scanner](https://aguarascan.com/): Static analysis of AI agent skills and MCP server configurations — 148+ rules, 13 threat categories, 3 detection layers (pattern matching, NLP analysis, taint tracking) - [Auto-Discovery](https://aguarascan.com/blog/how-i-built-semgrep-for-ai-agents/): Automatically discovers and scans MCP server configurations from Claude Desktop, Cursor, Windsurf, and other MCP clients - [Aguara Watch Observatory](https://aguarascan.com/blog/the-security-flywheel/): Continuous crawling and scanning of 42,655+ skills across 7 registries — live threat intelligence for the MCP ecosystem - [MCP Server](https://aguarascan.com/): Native MCP tool that gives AI agents direct access to scan, lookup, and threat intelligence capabilities - [CI/CD Integration](https://aguarascan.com/): SARIF output for GitHub Code Scanning, GitLab SAST, and CI pipeline integration ## Installation ``` go install github.com/aguarascan/aguara@latest ``` ## Blog Posts - [CVEs in Anthropic's Own MCP Servers: When Reference Implementations Teach the Ecosystem to Be Insecure](https://aguarascan.com/blog/anthropic-mcp-server-cves/) - [Mapping the Agentic AI Attack Surface: How Aguara Detects the Threats Researchers Identified](https://aguarascan.com/blog/agentic-ai-attack-surface-detection-mapping/) - [Aguara v0.4.0, MCP v0.3.0 & Watch Expansion — Coordinated Release](https://aguarascan.com/blog/v0-4-0-coordinated-release/) - [Securing Your OpenClaw Setup: 7 Checks + Automated Scanning](https://aguarascan.com/blog/securing-openclaw-setup/) - [Kali Linux + Claude Desktop: When Offensive Security Meets MCP](https://aguarascan.com/blog/