How yxsh-codes/multi-agent-ai-research-system works
SwarmScholar takes a research topic through web search, page scraping, report drafting, and critique. Users can run this flow through the Streamlit interface or the command-line pipeline; both invoke agent and chain components defined in agents.py. The search and reader agents are configured with tools.py functions, which connect to Tavily and fetch web pages. The sampled code supports these main calls and data handoffs; no README was provided, and app.py was only partially excerpted.
Main entry points: app.py Streamlit UI · pipeline.py CLI Pipeline · agents.py Search Agent · tools.py Web Search
Components 12
User Entry
- Streamlit UI
app.py - CLI Pipeline
pipeline.py
Research Workflow
- Search Agent
agents.py - Reader Agent
agents.py - Writer Chain
agents.py - Critic Chain
agents.py
Models And Tools
- Web Search
tools.py - URL Scraper
tools.py - Groq LLM
- Tavily Search
- Web Pages
Other components
- Researcher
Connections 18
- Researcher → Streamlit UI: enters topic
app.py - Researcher → CLI Pipeline: enters topic
pipeline.py - Streamlit UI → Search Agent: invokes
app.py - Streamlit UI → Reader Agent: invokes
app.py - Streamlit UI → Writer Chain: invokes
app.py - Streamlit UI → Critic Chain: invokes
app.py - CLI Pipeline → Search Agent: invokes
pipeline.py - CLI Pipeline → Reader Agent: invokes
pipeline.py - CLI Pipeline → Writer Chain: invokes
pipeline.py - CLI Pipeline → Critic Chain: invokes
pipeline.py - Search Agent → Web Search: registers tool
agents.py - Reader Agent → URL Scraper: registers tool
agents.py - Web Search → Tavily Search: queries
tools.py - URL Scraper → Web Pages: fetches page
tools.py - Search Agent → Groq LLM: uses model
agents.py - Reader Agent → Groq LLM: uses model
agents.py - Writer Chain → Groq LLM: uses model
agents.py - Critic Chain → Groq LLM: uses model
agents.py