release-skills
Pass
Audited by Gen Agent Trust Hub on Mar 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill performs network operations by pushing code and tags to a remote git origin. Evidence: Step 9 executes git push commands. Context: These operations are directed to the repository's configured remote, which is a standard and expected behavior for release management.
- [INDIRECT_PROMPT_INJECTION]: The skill handles untrusted data derived from the git commit history and PR metadata. Ingestion points: git log output and GitHub PR author names. Boundary markers: No explicit delimiters are specified for separating log data from command logic. Capability inventory: Includes file modification, shell command execution, and network access. Sanitization: No specific sanitization processes are mentioned for the text extracted from commit history.
Audit Metadata