How paradiddle131/dynamic-research-agent works
Project type and purpose
• This is a backend-only Python microservice exposing a versioned REST API (likely FastAPI, given the “app/main.py”, “uv.lock” and typical structure).
• Its purpose appears to be to scrape data, process it with Large Language Models (LLM), dynamically build schemas/models, and serve results via an HTTP endpoint (e.g., “research” endpoint).
File structure
• Top-level: configuration files (.env, pyproject.toml, requirements.txt), main entrypoint (app/main.py), environment lock (uv.lock).
• app/api/v1/endpoints: defines HTTP routes (research.py).
• app/api/v1/schemas: defines request payload schemas.
• app/core: houses the business logic layers:
• config.py – loads environment/configuration values
• scraper.py – fetches or scrapes external data
• llm.py – wraps calls to external LLM services
• dynamic_models.py – builds or modifies data models at runtime
• app/utils: cross-cutting concerns (logging_config.py).
Clues for architectural insights
• Versioned API (v1) suggests future extensibility.
• Separation into API, Core, and Utils layers shows clear layering.
• No database folder/file – likely stateless or uses external services (LLM API, web scraping).
Elements
• HTTP request/response
• Data flow between core services
• Outgoing calls to external systems
• Core services in another
• External services in a third
• Utilities (logging) as a sidebar or overlay
Guidelines summary:
• A “Logging” overlay or sidebar capturing cross-layer events.