dev-workflow

Pass

Audited by Gen Agent Trust Hub on Jul 2, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to run standard development commands such as npm install, npm run build, and npx tools (lint-staged, commitlint) within the project directory for code validation and building.
  • [EXTERNAL_DOWNLOADS]: Project templates reference external libraries through well-known registries like registry.npmmirror.com to resolve standard Node.js dependencies.
  • [PROMPT_INJECTION]: The skill features an ingestion surface for external content from Lark (flybook) and web pages. This potential surface for indirect prompt injection is strongly mitigated by a mandatory human-in-the-loop (HITL) confirmation process enforced at every transition between workflow phases using the question tool.
  • [SAFE]: Template code includes standard authentication patterns for frontend applications, such as managing tokens in localStorage and using Authorization headers. No hardcoded credentials or sensitive data exfiltration patterns were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 2, 2026, 07:46 AM
Security Audit — agent-trust-hub — dev-workflow