release-prep
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes several shell commands including
git,gh,go, andmise. These are used to verify repository status, run unit tests, and automate the release lifecycle through custom repository tasks. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from commit history and changelog fragments during the curation process.
- Ingestion points: Step 3 instructions require the agent to read deleted source fragments from
HEADand the repository's commit history to verify and curate release notes. - Boundary markers: The instructions do not provide explicit boundary markers or isolation protocols (such as XML tags or triple backticks with warnings) to prevent instructions embedded in commit messages or fragments from influencing the agent's behavior.
- Capability inventory: The skill has permissions to write to the local filesystem (editing
.mdentries), execute arbitrary shell commands viamise, and interact with the remote repository via the GitHub CLI. - Sanitization: There is no mention of sanitization, filtering, or escaping of the content ingested from commit messages or fragments before it is processed by the agent.
Audit Metadata