create-pr
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security risks were identified in the skill's instructions or scripts. The workflow is designed with a strong focus on user consent and data safety.
- [DATA_EXPOSURE]: The skill proactively addresses data exposure risks by enforcing strict rules (Rule 3.6) that prohibit internal identifiers, customer data, and internal URLs in PRs created for public repositories. It also includes a mandatory warning for users before pushing to public repos.
- [INDIRECT_PROMPT_INJECTION]: Although the skill interacts with external data (GitHub comments and CI status) through background watchers, the risk is mitigated by core rules that prioritize human session history and require explicit user confirmation for key actions like PR creation, preventing autonomous obedience to external content.
- [COMMAND_EXECUTION]: Shell command execution is limited to standard repository management tools (git, gh). The skill uses safe interpolation methods, such as quoted heredocs, to ensure that user-supplied text cannot lead to command injection during PR creation or updates.
Audit Metadata