Innovation Lab Examples

Fetch.ai Innovation Lab Examples

80+ production-ready AI agent examples in Python

Build autonomous AI agents, multi-agent systems and agentic AI workflows with uAgents, ASI:One, Agentverse, MCP, the A2A protocol, LangChain, CrewAI, Gemini, Claude and OpenAI.

Quickstart · Examples · Structure · Docker · Contributing · FAQ

📈 Star history

If these examples save you time, a ⭐ helps other developers find them.

Every folder here is a self-contained, runnable AI agent project with its own README, dependencies and environment template. Whether you are writing your first autonomous agent or architecting a multi-agent system with real payments, there is a working example to start from.

🎯 Who is this for?

• Beginners exploring autonomous agents and Fetch.ai for the first time
• Builders integrating LLMs, payments or Web3 into agent workflows
• Hackathon participants who need a working starter in minutes
• Contributors who want to share their agent examples with the community

⚡ Quickstart

Run your first agent in under two minutes:

# 1. Clone the repo
git clone https://github.com/fetchai/innovation-lab-examples.git
cd innovation-lab-examples

# 2. Pick an example (e.g. the hackathon quickstarter)
cd fetch-hackathon-quickstarter

# 3. Create a virtual environment and install dependencies
python3 -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt

# 4. Set up environment variables
cp .env.example .env
# Edit .env with your API keys

# 5. Run the agent
python agents/alice/agent.py

Or use the automated setup script from the repo root:

./setup.sh fetch-hackathon-quickstarter

Prerequisites: Python 3.10+, pip and git. Some examples require API keys (ASI:One, OpenAI, Stripe, etc.) — check each example's .env.example.

📁 Repository structure

innovation-lab-examples/
│
├── 📄 README.md                 # This file
├── 📄 CONTRIBUTING.md           # How to contribute
├── 📄 SECURITY.md               # Vulnerability reporting
├── 📄 ISSUES_GUIDE.md           # How to file a good issue
├── 📄 LICENSE                   # Apache 2.0
├── 🔧 setup.sh                  # One-command example setup
├── 🐳 Dockerfile                # Run any example in a container
├── 🐳 docker-compose.yml        # Compose support
├── 🔧 ruff.toml                 # Repo-wide lint baseline
│
├── 📂 .github/                  # CI and contributor templates
│   ├── workflows/               #   ci · pull_request_ci · review-required · pr-ai-review
│   ├── scripts/                 #   run-example-tests.sh (per-example test runner)
│   ├── ISSUE_TEMPLATE/
│   ├── pull_request_template.md
│   ├── CODEOWNERS
│   └── BRANCH_PROTECTION.md
│
├── 📂 scripts/                  # Repository tooling
│   └── pr-ai-review.mjs         #   ASI:One code review for pull requests
│
├── 📂 docs/                     # Shared guides and templates
│   └── AGENT_README_TEMPLATE.md
│
├── 📂 contributors/             # 👈 Community agents go here
│   ├── README.md                #   Contributor guide
│   ├── CHANGELOG.md             #   Community changelog
│   ├── community_agent/
│   ├── news-summarizer-agent/
│   └── gemini-task-manager-agent/
│
├── 🟢 Getting started
│   ├── fetch-hackathon-quickstarter/   # 👈 Start here
│   ├── av-script-example/
│   ├── asi-cloud-agent/
│   ├── deploy-agent-on-av/
│   ├── openclaw/
│   └── cursor-rules/
│
├── 🤖 LLM integration
│   ├── asi1-llm-example/
│   ├── dynamic-ui-on-asi1/          # ASI:One interactive cards (TripMate + news)
│   ├── anthropic-quickstart/
│   ├── gemini-quickstart/
│   ├── openai-agent-sdk/
│   ├── Claude Agent SDK/
│   ├── google-genai-parallel-processing/
│   ├── flight-tracker-openai-workflow-agent/
│   ├── langchain-agents/
│   ├── pydantic-agent/
│   └── security-scanner-agent/
│
├── 🔗 Agent-to-agent (A2A)
│   ├── launch-your-a2a-agent/
│   ├── launch-your-a2a-research-team/
│   ├── a2a-cart-store/
│   └── a2a-uAgents-Integration/
│
├── 🧩 MCP
│   └── mcp-agents/
│
├── 💰 Payments
│   ├── fet-example/
│   ├── image-agent-payment-protocol/
│   ├── stripe-horoscope-agent/
│   └── stripe-payment-agents/
│
├── 🧠 RAG and knowledge
│   ├── Rag-agent/
│   ├── llama-index/
│   └── pdf-summariser-example/
│
├── 👥 Multi-agent systems
│   ├── google-adk/
│   ├── Crewai-agents/
│   ├── ag2-agents/
│   └── video-to-map-agent/
│
├── 🌐 Web3
│   ├── web3/
│   └── duffel-agent/
│
└── 🔌 External integrations
    ├── Composio/
    ├── Browser-based-agents/
    └── frontend-integration/

📚 Examples index

🟢 Getting Started

🤖 LLM Integration

🔗 Agent-to-Agent (A2A)

🧩 MCP (Model Context Protocol)

💰 Payments

🧠 RAG & Knowledge

👥 Multi-Agent Systems

🌍 Community Contributors

New community agents belong in contributors/ — see contributors/README.md.

🌐 Web3 & Blockchain

🔌 External Integrations

🐳 Docker support

Run any example in a container without installing Python locally:

# Build and run a specific example
docker build --build-arg EXAMPLE=fetch-hackathon-quickstarter -t fetch-example .

# Run with your environment variables
docker run --env-file fetch-hackathon-quickstarter/.env fetch-example

Or use Docker Compose:

EXAMPLE=fetch-hackathon-quickstarter docker compose up

Several examples also include their own Dockerfile and docker-compose.yml for custom setups.

🤝 Contributing

We welcome contributions from everyone — a new agent example, a bug fix or a documentation improvement.

1. Star this repository (required before opening a PR)
2. Fork and create a feature branch from main
3. New agents go in contributors/<your-agent-name>/ — see contributors/README.md
4. Pick an issue — good first issues or feature challenges (real-time booking, payments, etc.)
5. Run linting — ruff check . && ruff format .
6. Run the tests — bash .github/scripts/run-example-tests.sh <your-example>
7. Open a PR using the PR template — PRs require maintainer review before merge

Every example should include: README.md, requirements.txt, .env.example (if env vars are needed), and a demo screenshot.

Use the Agent README Template for new examples.

📖 Resources

❓ FAQ

An AI agent is a program that perceives its environment, decides what to do and acts on its own to reach a goal — rather than waiting for step-by-step instructions. The agents in this repo use an LLM for reasoning and the uAgents framework for messaging, discovery and payments.

fetch-hackathon-quickstarter — it shows the orchestrator + worker pattern that most other examples build on. From there, pick the category matching your use case in the examples index.

Most examples need at least one. Each example ships a .env.example listing exactly what it reads. An ASI:One key covers the majority; individual examples may also need OpenAI, Anthropic, Google, Stripe or a vendor key.

Yes. Everything here is Apache 2.0 licensed. Third-party APIs each example calls have their own terms and pricing.

See deploy-agent-on-av for Agentverse deployment via Render, and asi-cloud-agent for ASI Cloud. Any example can also run in Docker — see Docker support.

Two patterns. A2A (agent-to-agent) connects independent agents over a shared protocol — see a2a-uAgents-Integration. Framework orchestration runs a team inside one process — see Crewai-agents, ag2-agents and google-adk.

Yes, and we would like you to. Put it in contributors/<your-agent-name>/ and read CONTRIBUTING.md first.

📄 License

Licensed under the Apache License 2.0.

Topics: ai-agents · autonomous-agents · multi-agent-systems · agentic-ai · llm-agents · uagents · fetchai · agentverse · asi-one · mcp · a2a-protocol · langchain · crewai · openai · python

Built by the Fetch.ai Innovation Lab and its contributors.