loop-on-ci
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to read and interpret external data from CI environments which may contain malicious instructions designed to influence the agent's automated fixing behavior.
- Ingestion points: The skill consumes external data through
gh pr checksandgh run view --log-failedcommands as specified in SKILL.md. - Boundary markers: The instructions do not define clear boundaries or 'ignore' directives to help the agent distinguish between informational log data and malicious commands embedded within logs.
- Capability inventory: To achieve its goal of fixing failures, the skill relies on the agent's ability to modify source code and perform git operations, providing a high-privilege execution path for successfully injected prompts.
- Sanitization: No sanitization, validation, or structural parsing is applied to the raw logs before the agent processes them.
Audit Metadata