feedback
Pass
Audited by Gen Agent Trust Hub on Apr 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the GitHub CLI (
gh) to automate the creation of issues on theVGrss/Acumenrepository, which is an expected vendor resource. - [PROMPT_INJECTION]: The skill processes free-form user input and interpolates it into a shell command, creating an attack surface for indirect prompt injection.
- Ingestion points: User-provided text for 'problem', 'impact', and 'solution' in SKILL.md.
- Boundary markers: Step 5 provides a review and confirmation phase before the command is executed, though explicit sanitization rules are missing.
- Capability inventory: Shell command execution via
gh issue createin SKILL.md Step 6. - Sanitization: Absent; the skill relies on the agent's default handling of command arguments.
Audit Metadata