pydantic-ai-researcher

What is pydantic-ai-researcher?

pydantic-ai-researcher is a research system that orchestrates an asynchronous loop between two specialized agents: a research agent and an evaluator agent. The research agent answers complex queries using external MCP servers, while the evaluator agent assesses and refines these answers, iterating until a satisfactory response is attained. It is designed for users who need automated, iterative research refinement.

How to use pydantic-ai-researcher?

After installing Python 3.x, use Pipenv (pipenv install) or pip (pip install -r requirements.txt) to install dependencies. Create a .env file with OPENAI_API_KEY, TAVILY_API_KEY, and BRAVE_API_KEY. Run python main.py and type research queries into the interactive prompt. To exit, type exit.

Key features of pydantic-ai-researcher

• Asynchronous research loop for iterative evaluation and improvement
• Dual agent system (research and evaluator agents)
• Pluggable MCP servers: Tavily, Brave Search, arXiv
• Environment configurable via .env files
• Straightforward installation with Pipfile or requirements.txt

Use cases of pydantic-ai-researcher

• Conduct iterative research on complex queries with automated feedback
• Refine answers through multiple evaluation-and-improvement cycles
• Gather data from diverse sources (web search, academic papers) via MCP servers
• Automate back-and-forth improvement of research answers until satisfactory

FAQ from pydantic-ai-researcher

What dependencies are required to use pydantic-ai-researcher?

Python 3.x is required. API keys for OpenAI, Tavily, and Brave Search must be set in a .env file.

How do I configure pydantic-ai-researcher?

Create a .env file at the project root and define OPENAI_API_KEY, TAVILY_API_KEY, and BRAVE_API_KEY. Other API endpoints may be defined as needed.

How does the pydantic-ai-researcher architecture work?

It implements a two-agent evaluator-optimizer loop. The research agent generates initial answers using external MCP servers, and the evaluator agent provides feedback for improvement. The loop iterates until the answer is accepted or the maximum iteration limit is reached.

How do I run pydantic-ai-researcher?

Execute python main.py from the project root. Type your research queries at the interactive prompt.

How do I exit the pydantic-ai-researcher program?

Type exit in the interactive prompt to stop the program.