superteam
Pass
Audited by Gen Agent Trust Hub on May 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
Bashtool to executegitandgh(GitHub CLI) commands for repository operations, including branch switching, issue retrieval, and PR management. These operations are restricted to the context of the active development task. - [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection due to its core function of ingesting data from GitHub issues and repository-hosted design artifacts.
- Ingestion points: Reads issue content via
gh issue viewand processes design/plan documents fromdocs/superpowers/. - Boundary markers: Uses structured artifact templates and mandatory Acceptance Criteria (AC) IDs to maintain context, though it lacks explicit technical delimiters for raw external input.
- Capability inventory: Access to
Bash(Git/GH CLI),Write, andEdittools across the repository. - Sanitization: Features a multi-layered defense including a mandatory 'Adversarial Design Review' phase and a 'Forbidden-append denylist (LC5)' in the project delta logic to prevent overrides from weakening core security rules.
Audit Metadata