refactorlib-cc
Pass
Audited by Gen Agent Trust Hub on Apr 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes dynamic context injection (the '!' syntax) in 'SKILL.md' to execute local shell commands such as 'bun -e' and 'ls' at load time to gather environment data and dependency information.
- [REMOTE_CODE_EXECUTION]: During the triage phase in 'Phase 3', the skill performs dynamic code execution (Category 10) of local source code and library equivalents using 'bun -e' to verify behavioral parity between handcrafted utilities and their replacements.
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) because it processes untrusted data from the audited project's files. • Ingestion points: 'package.json' and project source code. • Boundary markers: Absent. • Capability inventory: 'Bash', 'Agent', 'WebFetch', 'WebSearch', and 'mcp__bun__run-bun-eval'. • Sanitization: No sanitization of external file content is performed before processing.
Audit Metadata