news-aggregator-skill

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • EXTERNAL_DOWNLOADS (MEDIUM): The skill directs users to install code from an untrusted GitHub repository (cclank/news-aggregator-skill) via git clone and npx. This source is not on the trusted organizations list.
  • COMMAND_EXECUTION (MEDIUM): The agent is instructed to execute a shell command (python3 scripts/fetch_news.py) with arguments like --keyword that are derived from user input. While the instructions suggest the agent expand these keywords, there is a risk of command injection if the input is not correctly escaped before being passed to the shell.
  • PROMPT_INJECTION (LOW): The skill is vulnerable to Indirect Prompt Injection (Category 8).
  • Ingestion points: The fetch_news.py script with the --deep flag fetches the main text content from external web pages (e.g., Hacker News, 36Kr, Weibo).
  • Boundary markers: No boundary markers or instructions to ignore embedded commands are specified in the prompt interpolation logic.
  • Capability inventory: The skill has the capability to execute subprocesses (python3) and write files to the reports/ directory.
  • Sanitization: There is no evidence of content sanitization or validation before the fetched data is presented to the LLM for analysis.
  • DATA_EXFILTRATION (SAFE): While the skill makes network requests, they are directed at known news aggregators and social media platforms for the purpose of fetching news, consistent with the skill's primary function.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 01:55 PM