maixpy-dev

Pass

Audited by Gen Agent Trust Hub on Jul 18, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill acts as a reference for the MaixPy v4 SDK, providing functional examples for hardware interaction, image processing, and AI model deployment. All components, including network communication and file system operations, align with the intended use cases of an embedded development platform.\n- [PROMPT_INJECTION]: The skill documents integration patterns for Large Language Models (LLM) and Vision Language Models (VLM) which ingest potentially untrusted data, establishing an indirect prompt injection surface.\n
  • Ingestion points: Text input is processed via qwen.send() and visual data via vlm.ask() in the references/llm_vlm.md documentation.\n
  • Boundary markers: The code templates provided do not include delimiters or specific instructions to ignore embedded commands within the processed data.\n
  • Capability inventory: The execution environment includes network access (maix.http, maix.mqtt), hardware peripheral control (maix.uart, maix.gpio), and file system access to /root/.\n
  • Sanitization: Input validation and sanitization are not featured in the basic integration examples provided in the documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 18, 2026, 08:10 AM
Security Audit — agent-trust-hub — maixpy-dev