Release
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill makes extensive use of command-line tools to perform its tasks. It utilizes
jj(Jujutsu) for version control operations,gitfor pushing tags,denofor project quality checks (linting, formatting, testing), and specialized tools likesvbumpandchangelogfor release management. These commands are integral to the release process and operate within the project scope. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data in the form of commit messages and user-provided descriptions to generate changelog entries.
- Ingestion points: Commit messages are retrieved via
jj login Step 1, and user descriptions are accepted in Step 2. - Boundary markers: The
changelog addcommand uses double quotes to encapsulate the description, providing basic structural separation. - Capability inventory: The skill can execute shell commands, modify project configuration files (like
deno.json), and perform remote network operations viagit pushandjj git push. - Sanitization: No explicit sanitization or filtering of commit message content is specified in the instructions, though a manual user review step (Step 3) is included to verify the changelog before publication.
Audit Metadata