global-ai-news-brief

Pass

Audited by Gen Agent Trust Hub on Aug 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to run the included Python script scripts/fetch_all.py to collect data from various platforms. It also uses the shell open command to display the generated HTML report to the user.
  • [EXTERNAL_DOWNLOADS]: The fetch_all.py script makes network requests to https://redfox.hk to retrieve content. This domain belongs to the skill's author and is used for its primary data-gathering purpose.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes untrusted content from multiple social media platforms.
  • Ingestion points: The agent processes data from platforms such as TikTok, X, Douyin, and others via the results of scripts/fetch_all.py.
  • Boundary markers: The prompt lacks explicit delimiters or instructions to treat the ingested content as untrusted data or to ignore embedded commands.
  • Capability inventory: The agent has the ability to execute local scripts, write to the filesystem (specifically the output/ directory), and execute shell commands (open).
  • Sanitization: No specific filtering or escaping of the ingested content is defined before the AI agent performs its analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 20, 2026, 05:36 AM
Security Audit — agent-trust-hub — global-ai-news-brief