refactor
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill identifies and executes project-specific shell commands for testing, type checking, and linting (e.g., 'npm test', 'bun test', 'pytest', 'tsc --noEmit') discovered from configuration files like package.json.
- [DYNAMIC_EXECUTION]: The skill utilizes a local helper script ('scripts/ast_grep_helper.py') and structural search tools to modify source code. It also dynamically generates a team configuration file to orchestrate sub-agents for parallel task execution.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted codebase content which is then processed by sub-agents, creating a surface for indirect prompt injection. This is relevant given the skill's access to shell execution and filesystem modification capabilities.
- Ingestion points: Code exploration agents in Phase 1 and configuration file discovery in Phase 3.
- Boundary markers: The instructions do not define explicit delimiters to isolate external code content from agent instructions within interpolated prompts.
- Capability inventory: Access to 'bash' for shell execution, 'edit' for file writes, and LSP tools for symbol manipulation.
- Sanitization: The skill relies on post-change verification via LSP diagnostics and project tests rather than input sanitization of analyzed code.
Audit Metadata