fix-ci
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources which could be used to influence agent behavior.
- Ingestion points: The skill reads failing job annotations and logs (Step 1.4) and repository instructions (Step 1.1).
- Boundary markers: The instructions do not specify any delimiters or safety warnings to distinguish between the content of the logs/instructions and the agent's internal logic.
- Capability inventory: The skill has the ability to execute local commands, create branches, push code, and open PRs (Step 3).
- Sanitization: There is no mention of sanitizing or escaping the data read from external CI logs before the agent processes it.
- [COMMAND_EXECUTION]: The skill is designed to execute commands derived from the repository environment and CI configurations.
- Evidence: Step 3 instructions include "Run the exact failed check locally where possible", which involves executing commands that are defined within the failing repository or CI pipeline, potentially leading to arbitrary command execution if the repository is malicious.
Audit Metadata