Overview
What is ResearchMCP?
ResearchMCP is a Model Context Protocol (MCP) service that provides internet research capabilities by integrating with Perplexity AI’s API. It exposes a single “research” tool for querying up-to-date information on a wide range of topics, with graceful error handling and degraded mode.
How to use ResearchMCP?
Set required environment variables (e.g., PERPLEXITY_API_KEY) in a .env file, then build and run the service using go build and ./ResearchMCP. The service exposes a tool named research that accepts a query string argument.
Key features of ResearchMCP
- Internet research through Perplexity AI
- Integration with MCP protocol for tool discovery and invocation
- Graceful error handling with degraded mode
- Configurable via environment variables
- Structured logging support
Use cases of ResearchMCP
- Asking complex, open‑ended questions that require up‑to‑date web knowledge
- Obtaining research reports with sourced information for analysis or content creation
- Integrating real‑time web search into any MCP‑compatible client or agent
FAQ from ResearchMCP
What dependencies or runtime are required?
Go 1.18 or later and a valid Perplexity AI API key are needed to build and run ResearchMCP.
How do I configure the service?
Configuration is done through environment variables: PERPLEXITY_API_KEY (required), and optional ones like PERPLEXITY_MODEL, PERPLEXITY_TIMEOUT, PERPLEXITY_MAX_RETRIES, PERPLEXITY_INITIAL_BACKOFF, and PERPLEXITY_MAX_BACKOFF.
What input does the research tool accept?
The tool requires a single query string (the research question or topic). It returns a text response that includes source URLs.
How are API errors handled?
The service includes graceful error handling with a degraded mode, configurable retries (default 2), and exponential backoff.
Is there authentication support?
ResearchMCP uses a Perplexity API key supplied via the PERPLEXITY_API_KEY environment variable. No other authentication or transport layer is described.
