hap-mcp-app-builder

Pass

Audited by Gen Agent Trust Hub on Jun 26, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local Python script scan_apps.py to identify existing applications in the workspace and check the installation status.
  • [EXTERNAL_DOWNLOADS]: Connects to raw.githubusercontent.com to perform a version check against the author's official metadata file.
  • [REMOTE_CODE_EXECUTION]: Features a self-update logic that uses git pull or git clone to download and apply code updates from the vendor's GitHub repository (mingdaocom/hap-skills).
  • [PROMPT_INJECTION]: The skill is designed to ingest user requirements to generate structured application plans, presenting an indirect prompt injection surface.
  • Ingestion points: User-provided descriptions of business logic or system requirements.
  • Boundary markers: None explicitly defined in the instruction set for separating user input.
  • Capability inventory: Executes local Python scripts and writes structured JSON/Markdown files to the local project directory.
  • Sanitization: The skill relies on internal instruction-based guides and schemas to validate the generated application structure.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 26, 2026, 06:48 AM
Security Audit — agent-trust-hub — hap-mcp-app-builder