ruyi-project-walkthrough

Pass

Audited by Gen Agent Trust Hub on Jun 29, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill relies on local scripts (scripts/convert.sh, scripts/md_to_html.py, scripts/verify_sources.py, scripts/import_graph.py) to perform project analysis and file conversion. These are invoked via standard shell commands within the documented documentation workflow.
  • [DATA_EXPOSURE]: The generator ingests the user's project files to extract architecture and design patterns. To mitigate risk, it uses a 'Manifest-First' protocol which requires verifying every claim against the actual source code and recording it in a structured JSON log (sources-manifest.json).
  • [PROMPT_INJECTION]: No evidence of prompt injection patterns or instructions to bypass the agent's safety protocols were found. The skill focuses entirely on the technical task of generating documentation.
  • [EXTERNAL_DOWNLOADS]: The skill mentions installation via the standard Claude Code marketplace or via cloning from the author's official GitHub repository (github.com/zwyin/ruyi-skills). These are recognized well-known sources for developer tooling.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 29, 2026, 06:44 AM