krt-release-marshal
Pass
Audited by Gen Agent Trust Hub on May 12, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes standard version control and platform CLI tools (
git,gh) to manage repository state. Commands are derived from local metadata and executed according to a user-approved plan. - [DATA_EXFILTRATION]: Explicit safety rules are defined to prevent the accidental inclusion of secrets, tokens, or internal environment dumps in the generated pull request bodies.
- [PROMPT_INJECTION]: The skill implements a robust approval policy with distinct gates for local state changes and remote mutations, ensuring the agent cannot perform destructive or notification-causing actions without explicit user consent.
Audit Metadata