fix-ci
Pass
Audited by Gen Agent Trust Hub on Aug 22, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
ghCLI tool to fetch pull request check statuses, which is a standard and expected operation for CI management. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external CI logs and check links, creating a potential surface for indirect prompt injection.
- Ingestion points: Data enters the agent's context through PR check logs and external URLs identified in the
gh pr checksoutput. - Boundary markers: Absent. The instructions do not define delimiters or provide warnings to the agent to ignore embedded instructions within these external sources.
- Capability inventory: The agent is authorized to modify the local filesystem (applying fixes) and push changes to the repository, which could be exploited if the agent follows malicious instructions found in CI logs.
- Sanitization: Absent. There is no mention of filtering or validating the content retrieved from external logs or check links before it influences the agent's actions.
Audit Metadata