app-in-skill-creator

Pass

Audited by Gen Agent Trust Hub on Jul 11, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill framework utilizes local shell scripts (e.g., app/start.sh) and Node.js-based servers (Hono) to launch the operator UI and workflow execution helpers. This is the intended design for the App-in-Skill architecture and is restricted to local network access.
  • [EXTERNAL_DOWNLOADS]: The skill documentation references and encourages the use of established, well-known Node.js packages such as hono and @hono/node-server, and integrates with well-known service providers like Busabase for cloud-backed storage.
  • [SAFE]: The skill provides extensive, high-quality security instructions for developers, including:
  • [Onboarding]: Mandatory first-run setup gates to prevent execution with incomplete or unsafe configurations.
  • [Secret Management]: Explicit rules against hardcoding secrets or pasting them into chat, instead directing users to .env files or secure vaults.
  • [Data Access]: Implementation of a provider-based abstraction layer to ensure data sanitization and prevent direct file system manipulation.
  • [Locking]: A standard file-locking mechanism (agent.lock) to maintain state integrity during concurrent agent/user operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 11, 2026, 10:37 AM
Security Audit — agent-trust-hub — app-in-skill-creator