AgentSafe-AI/tooltrust-scanner
View on GitHub ↗Security scanner for AI agent tool definitions
19 ★6 forksGoUpdated 18d ago
What you need to know
Static security scanner for MCP tool definitions that assigns trust grades (A-F) and flags risky patterns, usable as MCP, CLI, or CI.
Install
go install or download the binary
Usage
- •Point the scanner at MCP tool definitions to get security grades and findings
Key features
- ✓Static analysis of MCP tool definitions
- ✓Trust grades A through F
- ✓Risk pattern detection
- ✓CLI, MCP, and CI usage
- ✓Security-focused
Best for
Security review of MCP tool definitions
Platforms: macOS · Linux · WindowsClients: MCP-compatible clients
Reviewed 2026-08-11
Topics
agent-safetyai-securityai-security-toolgatewaygolangmcpmcp-servermcp-toolsmodel-context-protocolprompt-injectionsecurity-scannersupply-chain-security
- Stars
- 19★
- Forks
- 6
- Language
- Go
- License
- MIT
- Created
- 2026-03-01
- Last push
- 2026-08-17