weibo-hot-search

Pass

Audited by Gen Agent Trust Hub on Jul 14, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill connects to the vendor-owned domain redfox.hk to fetch current Weibo trending data. This is a functional requirement for the skill to operate and involves a domain associated with the skill author.
  • [DATA_EXFILTRATION]: The REDFOX_API_KEY environment variable is transmitted to the vendor's API for authentication. This is standard behavior for API-based services and does not involve unauthorized exfiltration of user data.
  • [PROMPT_INJECTION]: The skill processes external trending titles, which constitutes a surface for indirect prompt injection. Ingestion points: Trending data retrieved from the redfox.hk API via the fetch_weibo_hot.py script. Boundary markers: None; the agent renders and analyzes the topic titles directly. Capability inventory: Execution of a local Python script and network access to the specific vendor API. Sanitization: The script performs basic numeric extraction for heat scores, though topic titles are processed as text. This surface is inherent to the skill's content-aggregation purpose and is considered low-risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 14, 2026, 07:41 AM
Security Audit — agent-trust-hub — weibo-hot-search