Research Assistant Team 🤖

A multi-agent research assistant built with Agno and powered by OpenAI models.
This project coordinates a team of specialist AI agents—each with dedicated tools—to perform tasks such as web searches, website crawling, YouTube analysis, GitHub exploration, Hacker News monitoring, email drafting, and general Q&A synthesis.

The agents collaborate under a Team Coordinator that delegates tasks, manages memory, and synthesizes results into clear, well-formatted answers.

---

🚀 Features

• Specialist Agents

  • 🌐 InternetSearcher – Web searches with DuckDuckGo
  • 🕸 WebCrawler – Extracts and summarizes website content
  • 📺 YouTubeAnalyst – Analyzes YouTube videos and captions
  • 📧 EmailAssistant – Drafts and sends professional emails via Resend
  • 💻 GitHubResearcher – Explores repositories, pull requests, and issues
  • 📰 HackerNewsMonitor – Tracks trending Hacker News stories
  • 🤝 GeneralAssistant – Synthesizes information and handles broad queries

• Team Coordinator

  • Delegates tasks intelligently to relevant agents
  • Synthesizes multi-agent outputs into a coherent final answer
  • Maintains session memory for contextual follow-ups
  • Handles retries with exponential backoff

• Streamlit UI

  • Chat interface for interacting with the assistant
  • Sidebar with debugging options (team memory & tool logs)
  • Session reset for fresh conversations

---

📂 Project Structure

.gitignore
.python-version
all_agents.py       # Defines specialist agents
create_team.py      # Builds the team of agents
main.py             # Streamlit app entry point
pyproject.toml      # Dependencies and project metadata

---

⚙️ Installation

1. Clone the Repository

git clone https://github.com/your-username/my-agents-crew.git
cd my-agents-crew

2. Create a Virtual Environment

python -m venv .venv
source .venv/bin/activate   # Linux/Mac
.venv\Scripts\activate      # Windows

3. Install Dependencies

pip install -e .

Dependencies include:

• agno – Multi-agent framework
• streamlit – Frontend for chat UI
• duckduckgo-search – Web search
• crawl4ai – Website crawling
• youtube-transcript-api – Video transcript extraction
• resend – Email sending
• PyGithub – GitHub API
• hackernews – Hacker News API
• python-dotenv – Environment variable management

---

🔑 Environment Variables

Create a .env file in the root directory:

OPENAI_API_KEY=your_openai_key
MODEL_NAME=gpt-4.1
RESEND_API_KEY=your_resend_key
EMAIL_FROM=your_email@example.com
EMAIL_TO=default_recipient@example.com
GITHUB_ACCESS_TOKEN=your_github_token

---

▶️ Usage

Run the Streamlit App

streamlit run main.py

Example Queries

• Web Search:
  “What are the latest AI breakthroughs?” → Uses InternetSearcher
• Crawl a Website:
  “Crawl https://example.com and summarize the homepage.” → Uses WebCrawler
• YouTube Analysis:
  “Summarize this video: https://youtu.be/dQw4w9WgXcQ” → Uses YouTubeAnalyst
• Send Email:
  “Draft an email to bob@example.com about our project.” → Uses EmailAssistant
• GitHub Research:
  “Find trending Python repositories this week.” → Uses GitHubResearcher
• Hacker News Trends:
  “What’s trending on Hacker News today?” → Uses HackerNewsMonitor
• Memory Check:
  “What was my first question to you?” → Tests memory persistence

---

🧠 Memory & Debugging

• Team Memory: Agents remember past interactions during a session.
• Sidebar Controls: Toggle memory view, enable tool logs, and reset the session.

---

🛠 Development Notes

• Python version: 3.13 (see .python-version)
• Use pyproject.toml for dependency management
• Memory is session-based and resets when the Streamlit app restarts or via the sidebar button

---

📜 License

MIT License – feel free to use, modify, and distribute.