reverse-design

Pass

Audited by Gen Agent Trust Hub on Nov 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes the git, gh (GitHub CLI), and npm binaries to retrieve source code, interact with pull requests, and manage its runtime dependencies. These operations are conducted within the local repository and against official GitHub API endpoints.
  • [EXTERNAL_DOWNLOADS]: The skill installs required Node.js libraries from the NPM registry and downloads a Chromium browser via Playwright for diagram rendering. It uses the --ignore-scripts flag during installation to mitigate risks from third-party package hooks. Note that some package versions in the manifest are ahead of official releases (e.g., typescript@6.0.3), which may indicate a placeholder or future-dated configuration.
  • [INDIRECT_PROMPT_INJECTION]: Because the skill ingests untrusted source code, it is susceptible to indirect prompt injection. It mitigates this by instructing the agent to treat analyzed code strictly as data rather than instructions and by implementing strict Mermaid diagram sanitization to prevent cross-site scripting (XSS).
  • [SAFE]: The skill demonstrates high security awareness through dedicated functions for credential redaction (masking API keys and secrets), strict path boundary validation to prevent directory traversal, and the use of trusted registries for external dependencies.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 18, 2026, 03:01 AM