Add Parallel AI Integration

To integrate Parallel AI web research, you configure the Micro-service Communication Protocol servers and update your agent runner scripts for seamless operation within the NanoClaw framework. This setup enables quick web searches and deep research tasks via distinct APIs.

In task automation, quick search performs fast web lookups using the Parallel Search API, while deep research conducts comprehensive analyses using the Parallel Task API. Deep research requires user permission and returns detailed reports later without blocking operations.

Asynchronous task execution works by leveraging the NanoClaw scheduler for non-blocking result polling. This allows your assistant to initiate deep research tasks and retrieve comprehensive analyses later without halting other ongoing operations.

Yes, you need an API key to use Parallel AI for web research. The integration requires API key configuration to authenticate and access the distinct APIs for both quick web searches and permission-based deep research tasks.

Yes, you can perform non-blocking web searches by integrating Parallel AI's MCP servers. The system uses the NanoClaw scheduler to handle asynchronous result polling, ensuring fast information retrieval without blocking the main assistant.

You should use deep research instead of a quick web search when you need comprehensive analysis and detailed reports. The assistant first performs a quick search, and if more depth is needed, it asks permission to use the deep research tool for in-depth analysis.