ai-project-seed
Pass
Audited by Gen Agent Trust Hub on May 22, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill facilitates the creation of a standard project skeleton and governance documentation. It uses local templates to generate markdown files such as AGENTS.md, BACKLOG.md, and docs/ARCHITECTURE.md.\n- [COMMAND_EXECUTION]: The skill includes a local PowerShell script (scripts/bootstrap_eval_workspace.ps1) used to automate the creation of a test environment for project seeds. The script uses standard cmdlets (New-Item, Set-Content) to create directories and basic JSON metadata files. No suspicious network activity or use of restricted system tools was identified.\n- [PROMPT_INJECTION]: Instructions in SKILL.md use structured logic and internal checks (Intent Gate, Quality Gate) to guide the agent, which helps ensure consistent behavior and reduces the risk of instruction override.
Audit Metadata