bootstrap-repository

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to run official platform generators (e.g., uv init, create-expo-app) and a bundled Python script (inspect_repository.py). These commands are standard for repository initialization and are used in a read-only or scaffold-only manner.
  • [EXTERNAL_DOWNLOADS]: The skill depends on well-known third-party generators and the GitHub CLI tool. These resources originate from well-known technology ecosystems (Apple, Expo, Python/Astral, GitHub) and are utilized according to their primary purpose.
  • [PROMPT_INJECTION]: The skill processes project metadata from local files (like package.json and GitHub workflows) and the GitHub API. While this represents a potential surface for indirect prompt injection from untrusted repositories, the skill follows best practices by using structured reporting and read-only inspection scripts.
  • Ingestion points: scripts/inspect_repository.py reads local repository configuration files and GitHub API responses.
  • Boundary markers: The Python script outputs structured JSON, and the skill provides a clear, segmented report structure for the agent to follow.
  • Capability inventory: The skill executes shell commands for scaffolding projects and reads repository metadata via standard APIs.
  • Sanitization: The inspection script validates GitHub repository names using regular expressions to ensure they conform to expected naming conventions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 07:39 PM
Security Audit — agent-trust-hub — bootstrap-repository