cursor-rules-engineer
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill uses standard, read-only shell commands (like grep, find, cat, and jq) to analyze local codebase metadata and source patterns. No network communication or remote code execution was detected.
- [DATA_EXFILTRATION]: The provided templates specifically include security rules that advise against hardcoding sensitive information like API keys and tokens, encouraging the use of environment variables instead.
- [PROMPT_INJECTION]: The skill features an indirect prompt injection surface as it ingests codebase data to generate instructions for AI agents. However, this is inherent to its primary purpose of codebase analysis and is managed through human review of the generated markdown files.
- Ingestion points: Codebase files (e.g., package.json, src/ directory) analyzed via shell commands.
- Boundary markers: Not explicitly defined in the rule templates.
- Capability inventory: Shell analysis commands (cat, grep, find, jq) and file generation.
- Sanitization: None detected, reliance is placed on the LLM's summarization and user verification.
Audit Metadata