gentle-ai-branch-pr
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill follows security best practices by requiring human-in-the-loop authorization for all potentially sensitive actions, including remote data reads and any mutations (commits, pushes, PR creation).
- [COMMAND_EXECUTION]: The skill utilizes standard development tools and commands, such as the GitHub CLI (
gh), Go testing tools (go test,go run), and local shell scripts (./docker-test.sh). These are used for intended repository management and verification tasks within the project context. - [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it ingests data from GitHub issues and pull requests. However, the instructions explicitly identify this data as untrusted and require human verification before acting on it, which significantly mitigates the risk.
- Ingestion points: GitHub issues (via
gh issue view) and Pull Requests (viagh pr view). - Boundary markers: Not explicitly defined for the data read, but the policy mandates human review of the content.
- Capability inventory: Local command execution (
go,gh, shell scripts). - Sanitization: None mentioned for tool output, but human-mediated action provides a verification gate.
Audit Metadata