cs-explore
Pass
Audited by Gen Agent Trust Hub on Jun 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is restricted to local repository operations, focusing on code analysis and documentation within the .codestable directory. No network-based exfiltration or credential access was detected.
- [COMMAND_EXECUTION]: The skill invokes a local script (.codestable/tools/search-yaml.py) to perform metadata-based searches on archived exploration documents. This internal tool usage is consistent with the skill's stated purpose and uses fixed paths within the repository environment.
- [PROMPT_INJECTION]: The skill has an indirect prompt injection surface because it processes untrusted data from the repository. 1. Ingestion points: Local source files and repository code read via glob/grep/read tools. 2. Boundary markers: None explicitly mentioned in the instructions for isolating code content. 3. Capability inventory: File system read/write access and local command execution (python). 4. Sanitization: Not explicitly implemented. The risk is minimized by the skill's 'Guard Rules' (守护规则) which require strict evidence-based documentation and prevent the AI from executing actions based on unverified information.
Audit Metadata