The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill retrieves and processes information from external and potentially untrusted sources, creating an attack surface for indirect prompt injection attacks. \n- Ingestion points: Untrusted data enters the agent's context through tools like ls_paper_fetch, ls_patent_fetch, ls_news_fetch, and the general web search tool as described in the Core Capabilities and Web Search Tool Usage sections. \n- Boundary markers: There are no instructions requiring the agent to use delimiters or ignore embedded commands when processing the fetched content. \n- Capability inventory: The skill is equipped with specialized tools for searching and fetching data from intellectual property, medicinal chemistry, R&D pipeline, and business development domains. \n- Sanitization: No mechanisms for content filtering, validation, or escaping are defined for external data prior to its inclusion in research reports.

precision-oncology