funding-signal-monitor

Pass

Audited by Gen Agent Trust Hub on Apr 10, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches real-time funding data from the public Hacker News Algolia API (hn.algolia.com).
  • [COMMAND_EXECUTION]: Orchestrates the search process by executing specialized scripts from related skills including twitter-mention-tracker, reddit-post-finder, and hacker-news-scraper.
  • [PROMPT_INJECTION]: The skill processes untrusted external content from social media and web forums, presenting a surface for indirect prompt injection.
  • Ingestion points: Aggregates unstructured data from Twitter, Reddit, Hacker News, and general web searches in Phase 2.
  • Boundary markers: The prompt instructions in Phase 3 and 4 do not specify delimiters (e.g., XML tags or triple backticks) to isolate the untrusted external content from the agent's instructions.
  • Capability inventory: Resulting data is used to generate outreach reports, write to Google Sheets, and can be passed to subsequent outreach skills.
  • Sanitization: The provided Python script filters for funding stages and dollar amounts but does not sanitize the input text for malicious instruction patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 10, 2026, 10:47 AM