no-comments
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes a local script
python3 <pstack-runtime>/scripts/roles.pyduring Step 1 to determine role assignments and delegation targets. This script is part of thepstack-runtimeenvironment referenced by the skill's instructions. - [INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection in its delegation process.
- Ingestion points: File contents and git diffs (untrusted data) are read from the repository and pasted into the task brief for the delegated "Comment Sicko" agent.
- Boundary markers: The skill lacks explicit delimiters, XML tags, or "ignore embedded instructions" warnings between the agent persona and the untrusted file content/diffs.
- Capability inventory: The delegated agent is authorized to edit files in the shared checkout and invoke secondary tools such as
/how,/why, and/architect. - Sanitization: No sanitization, escaping, or validation of the file content is performed before it is interpolated into the child agent's context.
Audit Metadata