Read the full article on DataCamp: LangManus - A Guide With Demo Project

Learn how to build a multi-agent system using LangManus to analyze a trending GitHub repository, scrape its commit history, visualize activity trends, and generate a markdown report.

Overview

LangManus is an open-source, community-driven AI automation framework designed for building structured, multi-agent pipelines powered by LLMs. It orchestrates tools like LangGraph, LiteLLM, and browser-based agents to perform tasks such as planning, research, scraping, code analysis, and reporting.

Project Overview: GitHub Repository Analyzer

We’ll build an interactive LangManus-powered assistant that:

  • Finds a trending open-source GitHub repo
  • Scrapes its commit activity
  • Analyzes feature updates and contributions
  • Visualizes trends through charts
  • Generates a markdown report

Structure

LangManus-GitHub-Demo/
├── main.py
├── agent.py
├── planner.py
├── streamlit_app.py
├── agents/
│   ├── researcher.py
│   ├── browser.py
│   ├── coder.py
│   └── reporter.py
├── commit_chart.png
├── category_chart.png
├── topics_chart.png

Step-by-Step Implementation

Step 1: Prerequisites

python3 --version  # Ensure Python 3.8+
pip install requests beautifulsoup4 matplotlib streamlit
export GITHUB_TOKEN=your_personal_token

You’ll need a GitHub token to avoid rate limits.

Step 2: Creating the Planner and Agent Controller

planner.py

def plan_task(user_query):
    return [
        {'agent': 'researcher', 'task': 'Find trending repo'},
        {'agent': 'browser', 'task': 'Scrape GitHub activity'},
        {'agent': 'coder', 'task': 'Analyze recent commits and features'},
        {'agent': 'reporter', 'task': 'Generate Markdown report'}
    ]

agent.py

class LangManusAgent:
    def __init__(self, task):
        ...
    def run(self):
        ...
    def run_and_return(self):
        ...

It uses the planner to execute a chain of agents with shared context.

Step 3: Implementing LangManus Agents

agents/researcher.py

Scrapes the trending GitHub Python repo using BeautifulSoup.

agents/browser.py

Fetches recent commit history via GitHub REST API and extracts messages, authors, and timestamps.

agents/coder.py

Analyzes the commit history and generates:

  • 📈 Commit frequency chart
  • 🧩 Commits by category
  • 🔥 Most frequent words in messages

agents/reporter.py

Generates a markdown report combining:

  • Repository link
  • Recent commit messages
  • Summary by commit type

Step 4: Streamlit UI

streamlit_app.py

if st.button("🔍 Run Analysis on Trending Repo"):
    ...
    st.markdown(report)
    for path in chart_paths:
        st.image(Image.open(path))

Streamlit dashboard to trigger agents and display the markdown + charts.

Step 5: Run the App

streamlit run streamlit_app.py

Demo Screenshots

  • 📋 Markdown report
  • 📊 Commit activity over 30 days
  • 🔍 Category and topic analysis
  • 🖼 Charts generated using matplotlib

Conclusion

In this project, we demonstrated the power of LangManus to orchestrate a chain of agents to perform real-world analysis on a trending GitHub repo.

We:

  • 🔎 Researched trending projects
  • 📤 Scraped commit data
  • 📊 Visualized contribution patterns
  • 📝 Generated structured markdown summaries

LangManus is ideal for building agents that research, analyze, and report from real-world data pipelines.

Explore the full tutorial on DataCamp
👉 Click here to read it