The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill researches information from external web sources via Perplexity tools, which introduces an attack surface where malicious content on the web could attempt to influence the agent's behavior. • Ingestion points: Data enters the context through search results and analysis generated by the mcp__perplexity__perplexity_search and mcp__perplexity__perplexity_research tools as described in SKILL.md. • Boundary markers: The skill provides instructions for structuring the output into markdown headers and bullet points but does not define explicit delimiters to separate untrusted tool output from the agent's own instructions during processing. • Capability inventory: The skill uses Perplexity MCP tools for information retrieval. No capabilities for file system modification, shell execution, or direct network requests are present. • Sanitization: The skill includes instructions to verify major claims against official company blogs and verified press releases, which serves as a logical validation step to mitigate the impact of false or malicious information.

researching-ai-news