Research-Assistant

This monorepo contains an implementation of LLM agents that help with research. It consists of the following parts:

• automator implements the agent as a python sdk built on top of MCP clients
• a number of MCP servers implement tools that can be used by our agent or in any other MCP client:
  • terminal-mcp implements tools for an interactive terminal capable of running background tasks, a jupyter notebook tool, and tools to interact with a local codebase
  • talk-to-model contains a send_message tool for agents that help evaluate other LLMs
  • web-mcp contains a google search and a markdown broswer tool, but it's currently not working so well: often, websites are turned into markdown documents that exceed the token limits
• squiggpy contains a squiggle-like python library and a prompt for agents to use the library

Setup

Run python install.py to setup everything up

Testing

The automator component includes pytest-based tests to verify core functionality:

Running Tests

# Navigate to automator directory
cd automator

# Install with dev dependencies
pip install -e .[dev]

# Run all tests
pytest

# Run specific test files
pytest tests/test_quickstart.py
pytest tests/test_rag_hook.py

# Run tests with verbose output
pytest -v

# Run tests excluding RAG functionality (if dependencies not available)
pytest -m "not rag"

Test Coverage

• test_quickstart.py: Tests basic agent creation and terminal tool functionality (based on examples/quickstart.py)
• test_rag_hook.py: Tests RAG (Retrieval-Augmented Generation) hook functionality (based on examples/rag/rag_hook_example.py)
• test_system.py: Tests core system functionality including imports, RAG operations, and workspace setup

Test Requirements

• Basic tests require only core dependencies
• RAG tests require the optional [rag] dependencies: pip install -e .[rag]
• All tests use temporary directories to avoid affecting your actual workspace