xyq-nest-skill

Pass

Audited by Gen Agent Trust Hub on Aug 31, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses Python's urllib to make HTTP requests to the platform's API. While the instructions suggest executing scripts via python3 {baseDir}/scripts/..., these are part of the skill's own distribution and perform legitimate integration tasks.
  • [EXTERNAL_DOWNLOADS]: The download_results.py script downloads generated media files from URLs provided by the API. It uses urllib.request.urlopen with a standard User-Agent. These downloads are triggered by the user's generation requests.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided messages and external API responses (status updates). It explicitly instructs the agent to act as a "transporter" and avoid modifying or interpreting the prompts locally, which reduces the risk of the agent being manipulated by malicious instructions embedded in data, though a surface for indirect injection exists when the agent displays API results to the user.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 31, 2026, 06:35 AM
Security Audit — agent-trust-hub — xyq-nest-skill