change-spec-planner
Pass
Audited by Gen Agent Trust Hub on Mar 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script
init_change_doc.pyto automate the generation of documentation files within thedocs/changes/directory. This script utilizes standard libraries to sanitize filenames (slugification) and perform file system writes. All operations are confined to the local project structure. - [PROMPT_INJECTION]: The skill processes untrusted user data (requirements, issues, bug descriptions) to populate technical specifications. This creates an indirect prompt injection surface where malicious instructions in the input could be incorporated into the generated
.mdfiles. This risk is mitigated by the skill's own 'Risk Matrix' and 'Review Rules,' which mandate human sign-off for major changes. - [SAFE]: No hardcoded credentials, network exfiltration, or obfuscation techniques were detected. The skill follows best practices by separating planning (documentation) from execution (code changes).
Audit Metadata