multi-content-feed

Pass

Audited by Gen Agent Trust Hub on Aug 11, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill makes network requests to doubaoya.com to fetch content trending data. As this is the official domain for the tool's provider ('zizhanovo'), it is considered a legitimate vendor resource.
  • [COMMAND_EXECUTION]: The skill uses a Python script to perform API calls via urllib. The script is executed via python3 with arguments for platforms, keywords, and time ranges. All inputs are parsed safely, and no shell injection vectors were identified.
  • [CREDENTIALS_UNSAFE]: The skill correctly instructs the user to store their API key in an environment variable (DOUBAOYA_API_KEY) rather than hardcoding it. The instructions explicitly forbid printing or echoing the key.
  • [DATA_EXFILTRATION]: Data retrieved from the API is printed to stdout for the agent to process. The script only communicates with the trusted vendor domain doubaoya.com and does not attempt to send local system data elsewhere.
  • [PROMPT_INJECTION]: The instructions are clear and focused on data retrieval and presentation. No attempts to override agent safety guidelines or system prompts were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 11, 2026, 07:48 PM
Security Audit — agent-trust-hub — multi-content-feed