geekmagic-smalltv-ultra

Pass

Audited by Gen Agent Trust Hub on Aug 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes curl to interact with the device's REST API and pip or pio (PlatformIO) for environment setup and firmware compilation. These commands are consistent with the skill's stated purpose for IoT device management.
  • [EXTERNAL_DOWNLOADS]: The skill downloads the PlatformIO development suite and clones community firmware repositories from GitHub. These are well-known resources within the ESP8266 development community.
  • [DATA_EXFILTRATION]: The skill identifies that the device's /config.json endpoint exposes sensitive WiFi credentials. It provides the command for access but explicitly warns the user of its sensitive nature.
  • [PROMPT_INJECTION]: The skill creates an indirect prompt injection surface by ingesting and persisting JSON data from the local IoT device into a local configuration file (.claude/geekmagic-smalltv-ultra.local.md).
  • Ingestion points: Device metadata endpoints (e.g., /v.json).
  • Boundary markers: Absent; the skill relies on the integrity of the user-specified local IP.
  • Capability inventory: Access to curl, pip, and pio for command execution.
  • Sanitization: The skill performs basic key-value extraction but lacks rigorous validation of the strings returned by the device.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 1, 2026, 09:06 PM
Security Audit — agent-trust-hub — geekmagic-smalltv-ultra