Practical AI agent implementations across multiple frameworks with specialized agents for file operations, research, Jira integration, and more.
- ✅ Butler_Basil - Generic task coordination
- ✅ FileOps_Freddy - File operations (98.9% success)
- ✅ Jira_Johnny - Jira integration (100% success)
- ✅ Scrum_Sam - Multi-agent Scrum Master
- ✅ Story_Sage - User story specialist
- ❌ Data_Daniel - Tool schema errors
- ❌ Stocks_Sarah - MCP timeout issues
- ✅ Complex_Coding_Clara - 14-agent coding system with specialists (Architect, Senior Coder, Python, Web, Database, DevOps, Data Science, Agile, etc.)
- ✅ DeepResearch_Dave - Comprehensive research agent
- ✅ QuickResearch_Quinten - Rapid research (30-90s responses)
Manual Installation:
# Install dependencies
pip install -r requirements.txt
# Setup environment
cp example_env .env # Add your API keysRun Agents:
# GoogleADK web interface
cd GoogleADK && adk web
# AWS Strands agents - Recommended: Use Basic-Agent-Chat-Loop
# https://github.com/Open-Agent-Tools/Basic-Agent-Chat-Loop
pip install basic-agent-chat-loop
chat_loop /path/to/agent-examples/AWS_Strands/Complex_Coding_Clara/agent.py
# Optional: Save aliases for quick access
chat_loop --save-alias clara /path/to/Complex_Coding_Clara/agent.py
chat_loop clara # Use the alias
# AWS Strands agents - Direct execution (basic)
cd AWS_Strands/Complex_Coding_Clara && python agent.pyAdd to .env file:
# Choose your model provider
ANTHROPIC_API_KEY=your_key_here
OPENAI_API_KEY=your_key_here
GOOGLE_API_KEY=your_key_here
# Optional: Web search
TAVILY_API_KEY=your_key_here
# Optional: Local models
OLLAMA_BASE_URL=http://localhost:11434- CLAUDE.md - Comprehensive development guide
- Individual agent READMEs in framework directories