release-skills

Pass

Audited by Gen Agent Trust Hub on Jun 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes a series of Git commands (git tag, git log, git diff, git add, git commit, git push) to automate the release process. These operations are restricted to the local environment and the configured remote repository.
  • [DATA_EXFILTRATION]: The skill includes functionality to push local commits and tags to a remote server via git push. This is a standard and expected feature for a release management tool and is guarded by a user confirmation step.
  • [PROMPT_INJECTION]: The instructions include a directive to follow the full release workflow even when a user requests a direct push, which acts as a process integrity check.
  • [PROMPT_INJECTION]: Indirect Prompt Injection Surface:
  • Ingestion points: Commit messages ingested via git log in Step 2 and Step 5.
  • Boundary markers: Absent for the ingestion of git log data.
  • Capability inventory: File writing (Step 7) and network operations (Step 9: git push).
  • Sanitization: Absent; the agent is tasked with summarizing and translating content from commit history directly into changelogs.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 28, 2026, 03:19 AM
Security Audit — agent-trust-hub — release-skills