crypto-research

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 1.00). The skill explicitly requires agents to fetch and ingest external web content (e.g., SKILL.md "Web Page Fetching" using WebFetch to https://markdown.new/... or https://r.jina.ai and multiple agent prompts that instruct WebSearch/news/social-media sentiment collection such as agent-prompts/crypto_coin_analyzer_agent_prompt.md and crypto_news_scanner_agent_prompt.md), and that untrusted third-party content is read and used to drive analysis and investment decisions (e.g., mandates to use 5+ sources and to research MINIMUM_SOURCES), which meets all criteria for exposure to potential indirect prompt injection.