architect-review

Pass

Audited by Gen Agent Trust Hub on May 3, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides numerous shell commands for auditing dependencies (npm audit), installing packages (npm install), and running code analysis tools (npx madge, npm ls).
  • [EXTERNAL_DOWNLOADS]: Fetches documentation and code transformation scripts from external repositories, specifically from the official Facebook/Meta and ReactJS organizations on GitHub (e.g., raw.githubusercontent.com/facebook/react/...).
  • [REMOTE_CODE_EXECUTION]: Documentation describes running jscodeshift with remote transformation URLs. This allows the execution of JavaScript code fetched from the network onto local source files.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it is designed to review and process user-provided code changes and architectural designs.
  • Ingestion points: Analyzes software designs and code changes for review (SKILL.md).
  • Boundary markers: No explicit instructions to ignore embedded commands in the data being reviewed.
  • Capability inventory: Includes capabilities for package installation, file system analysis, and remote script execution through provided guides.
  • Sanitization: No validation or sanitization of external content is specified before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
May 3, 2026, 06:21 PM
Security Audit — agent-trust-hub — architect-review