build-mcp-use-server
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFENO_CODEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [NO_CODE]: The skill consists of markdown documentation and instructional content for an AI agent. No executable scripts, binaries, or active code files were found in the analyzed file set.
- [EXTERNAL_DOWNLOADS]: The documentation guides the AI and user to install standard development packages from the npm registry (e.g., mcp-use, zod, react) and clones project templates from the mcp-use organization on GitHub. These references target well-known and trusted development resources.
- [INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to scan local project files (such as package.json, import statements, and server configuration) to identify the framework version and project state. This ingestion of potentially untrusted project data is a standard development workflow, but technically represents an indirect prompt injection surface.
- Ingestion points: SKILL.md (Scan what already exists step).
- Boundary markers: Absent.
- Capability inventory: Filesystem reading and code generation.
- Sanitization: Absent.
- [SAFE]: No malicious patterns, data exfiltration, or obfuscation techniques were detected. The skill promotes secure development practices, such as the use of environment variables for secrets and ensuring .env files are excluded from version control via .gitignore.
Audit Metadata