yeelight-cinema-director

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local shell and PowerShell scripts (scripts/invoke.sh, scripts/invoke.ps1) to manage the lifecycle of a Node.js background service and to interface with the yeelight-home CLI tool. These executions are confined to local service management and hardware control within the skill's intended scope.- [EXTERNAL_DOWNLOADS]: The skill fetches cinema and soundtrack metadata from established external services including IMDb, Apple iTunes, LRCLIB, and Google APIs. These interactions are performed via dedicated adapters that implement security best practices, such as DNS validation to prevent Server-Side Request Forgery (SSRF) against local or private network addresses.- [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from external media catalogs (film titles, metadata, and lyrics). It mitigates injection risks by enforcing strict length limits via boundedString and cleanQuery utilities, and by transmitting data to the agent in structured JSON format rather than executable markup.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 01:28 PM
Security Audit — agent-trust-hub — yeelight-cinema-director