sol-loop
Pass
Audited by Gen Agent Trust Hub on Nov 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the local environment and remote repositories, including git diffs and GitHub Pull Request metadata, which are then passed to a subagent for analysis. Instructions hidden within code comments or PR descriptions could potentially influence the behavior of the review subagent or the parent agent.
- Ingestion points: Data is retrieved via
git diff,git status, andgh pr view. - Boundary markers: The skill provides a structured prompt for the subagent but does not implement explicit delimiters or warnings to ignore instructions embedded within the code being reviewed.
- Capability inventory: The skill uses
gitandghCLI tools and utilizes platform-specific capabilities likedelegate_taskandorchestrator_capabilitiesto invoke other AI models. - Sanitization: There is no evidence of sanitization or escaping of the code diffs or PR data before they are interpolated into the review prompt.
- [COMMAND_EXECUTION]: The skill frequently executes shell commands to perform repository operations (
git status,git fetch,git merge-base,git diff) and GitHub interactions (gh pr view). These commands are used to identify the scope of changes and retrieve base branch information.
Audit Metadata