github-repo-architect

Pass

Audited by Gen Agent Trust Hub on Jul 13, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • [SAFE]: The skill operates as a high-level architectural planner. It does not perform network operations, access local sensitive files, or execute code.
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to include references to official, well-known GitHub Actions (such as actions/setup-python and actions/setup-node) when generating CI/CD workflows. These are standard industry tools for repository automation.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted user-provided project descriptions to generate content like configuration files and shell commands.
  • Ingestion points: Project ideas and requirements provided by the user in the prompt (SKILL.md).
  • Boundary markers: None explicitly specified for the output formatting of generated files.
  • Capability inventory: Limited to text generation; the skill itself does not have file system or network access.
  • Sanitization: Includes explicit guardrails forbidding the invention of secrets, API keys, or package names, and requires security documentation for sensitive use cases.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 13, 2026, 01:07 AM
Security Audit — agent-trust-hub — github-repo-architect