init-project
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFECOMMAND_EXECUTIONPERSISTENCEINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes the
project-init-labelsscript and the GitHub CLI (gh label create) to manage repository labels during the optional setup phase.\n- [PERSISTENCE]: The skill installs apre-commithook (copied fromtemplates/config/pre-commit) into the local repository, which automatically executes during the git commit workflow to run linting tools like Ruff.\n- [INDIRECT_PROMPT_INJECTION]: The skill generates persistent instruction files (AGENTS.md,CLAUDE.md, and the.rules/directory) that govern how AI agents interact with the project in future sessions.\n - Ingestion points: Template files are read from the
templates/directory and written to the project root.\n - Boundary markers: No explicit boundary markers are added to the generated documentation files.\n
- Capability inventory: The skill possesses file-writing capabilities and can execute shell commands (
gh,project-init-labels).\n - Sanitization: The skill performs variable interpolation for project metadata (e.g.,
{{PROJECT_NAME}}) using values detected from the existing environment.
Audit Metadata