skills/cristoslc/swain/swain-release/Gen Agent Trust Hub

swain-release

Pass

Audited by Gen Agent Trust Hub on May 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes several local scripts located in the .agents/bin/ directory of the repository (e.g., swain-session-check.sh, security-scan.sh, swain-trunk.sh, swain-bookmark.sh) to perform environment checks and release operations. It also uses standard git and grep commands to manage the release workflow.
  • [EXTERNAL_DOWNLOADS]: The changelog generation script scripts/render_changelog.py requires the well-known jinja2 Python package for rendering templates.
  • [PROMPT_INJECTION]: The skill incorporates an indirect prompt injection surface by processing git commit messages and a project-specific override file (.agents/release.override.skill.md). This risk is mitigated by explicit instructions for the agent to synthesize and bucket the data into a human-readable narrative rather than directly transcribing untrusted text.
Audit Metadata
Risk Level
SAFE
Analyzed
May 7, 2026, 09:46 PM