list-good-prs
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes a Bash script that executes
gh(GitHub CLI) andjqto query repository information and pull request metadata. - [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it fetches and reports pull request titles, branch names, and author information which are externally controlled by any user who can open a PR in the repository.
- Ingestion points: Untrusted data is ingested from the GitHub API via
gh pr listandgh pr viewcommands inSKILL.mdto populate a summary table. - Boundary markers: The skill does not employ boundary markers or specific instructions to prevent the agent from following directives that might be embedded in pull request titles or descriptions.
- Capability inventory: The environment allows for execution of shell commands (
bash,gh,jq) and network requests to GitHub's API. - Sanitization: The skill does not sanitize the text content of the pull requests before displaying it in the output table.
Audit Metadata