agent-deep-research

This skill equips AI agents with deep research and RAG-based document search capabilities, enabling them to perform multi-step, asynchronous analysis and synthesis of complex information to generate expert-level reports, rather than simply aggregating search results. It addresses the shortcomings of AI agents when handling tasks that require in-depth investigation, cross-document information integration, or analysis in specialized domains.

To use it, simply install this skill for your AI agent, configure your Google API key and the uv tool, and you can initiate research tasks via simple commands or by conversing with the agent. It supports using local files as research context and can persist the research process state, facilitating subsequent queries and report retrieval.

This skill is ideal for various users who need to process complex information, write in-depth reports, or conduct analysis in professional fields, such as researchers, analysts, content creators, students, and anyone who wants their AI assistant to 'think' rather than just 'look up'.

It is recommended to carefully review the configuration requirements before use to ensure environment variables are set correctly. For analyses involving a large number of local files, note that using the --context-extensions parameter to filter file types can improve efficiency and accuracy.