dev-pr
Pass
Audited by Gen Agent Trust Hub on May 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands using
gitand the GitHub CLI (gh) to manage branches, view pull request details, and post replies to comments. It also dynamically adapts to project-specific quality gate commands such aspnpm lintornpm testbased on the project's configuration files (e.g., package.json, composer.json).\n- [PROMPT_INJECTION]: The skill processes untrusted external data from GitHub PR comments and CI failure logs. This exposes an indirect prompt injection surface where a malicious actor could include instructions in a PR comment or CI log to influence the agent's actions during the feedback resolution phase.\n - Ingestion points: PR comments are fetched from the GitHub API and CI logs are retrieved via the
gh run viewcommand as described inSKILL.md.\n - Boundary markers: The skill does not define explicit delimiters or instructions to ignore embedded commands within the fetched external content.\n
- Capability inventory: The agent has permissions to edit local source code, create commits, and push changes to remote repositories.\n
- Sanitization: There is no evidence of sanitization, validation, or filtering of the comment body or log content before the agent processes and actions it.
Audit Metadata