release-notes
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a variety of shell commands using
gitandgh(GitHub CLI) to resolve repository paths, fetch tags, and retrieve commit logs and diffs. - Evidence:
git rev-parse --show-toplevel,git -C <dir> fetch origin --tags --prune,gh release view, andgit -C <dir> log --oneline --no-merges. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data (commit subjects, file contents, and diffs) which could contain instructions designed to override agent behavior during the classification and rendering stages.
- Ingestion points: Untrusted data enters the context through
git log,git diff, andgit show(SKILL.md sections 2, 3, and 4). - Boundary markers: The instructions do not specify any boundary markers (like XML tags or specific delimiters) to isolate the untrusted git output from the agent's instructions.
- Capability inventory: The skill has capabilities to execute shell commands and write to the chat context.
- Sanitization: There is no explicit sanitization or filtering mentioned for the content extracted from commit messages or diffs.
Audit Metadata