# Fetch.ai Innovation Lab Examples: 80+ AI Agent Starter Projects > fetchai/innovation-lab-examples is a Python collection of more than 80 self-contained agent examples covering uAgents, LangChain, CrewAI, Claude, Gemini, and OpenAI, each with its own README, dependencies, and environment template. The repository targets anyone from first-time agent builders to teams prototyping multi-agent systems with payments. **fetchai/innovation-lab-examples** — 80+ production-ready AI agent examples in Python — build autonomous agents, multi-agent systems and agentic AI with uAgents, ASI:One, MCP, A2A, LangChain, CrewAI, Gemini, Claude and OpenAI. - Repository: https://github.com/fetchai/innovation-lab-examples - Website: https://innovationlab.fetch.ai/resources/docs/intro - Stars: 1,144 · Forks: 84 - Language: Python - License: MIT - Published: 2026-09-10 · Updated: 2026-09-10 · Language: en - Canonical page: https://hysenlabs.com/projects/fetchai-innovation-lab-examples ## What the Repository Provides and Who It Is For Every folder in this repository is a self-contained, runnable agent project. The README describes each example as having its own README, dependencies, and environment template, so picking one up does not require reading the rest of the repository first. The README groups its intended users into four categories: beginners exploring autonomous agents for the first time, builders integrating LLMs, payments, or Web3 into agent workflows, hackathon participants who need a working starter in minutes, and contributors who want to share their own agent examples. The examples span a wide range of frameworks. The tech stacks listed in the examples index include uAgents, the Fetch.ai agent framework; ASI:One, the Fetch.ai LLM API; LangChain, CrewAI, Pydantic AI, the Anthropic Claude SDK, Google Gemini, and the OpenAI Agents SDK. Integration examples also cover MCP (Model Context Protocol), the A2A protocol for agent-to-agent communication, and payment protocols using Stripe and Skyfire. The repository is an open collection, not a framework. It does not provide shared utilities that examples call into. Each folder is independent. ## How the Repository Is Organized The repository root contains the setup.sh one-command helper, a Dockerfile, docker-compose.yml, ruff.toml for linting, and the contributing and security guides. Each example lives in its own directory with a predictable structure: a README, a requirements.txt, and a .env.example. The README organizes examples into categories. Getting Started covers the hackathon quickstarter template and basic Agentverse deployment. LLM Integration includes series for Claude (basic through multi-agent) and Gemini (text, Imagen, Veo, research, film), along with standalone examples for OpenAI, LangChain, and Pydantic AI. Agent-to-Agent covers A2A protocol examples at beginner and intermediate difficulty. Additional sections cover MCP agents, CrewAI, LlamaIndex, Composio, and Web3 integrations. Difficulty labels in the examples index use a three-level scale: beginner, intermediate, and advanced. The advanced examples combine multiple concerns: the langchain-agents example is described as a hackathon competitive-intelligence agent with Stripe payments, and the pydantic-agent example uses Pydantic AI with ASI:One interactive cards and the Shippo shipping API. A series entry like the anthropic-quickstart spans beginner through advanced across its sub-examples. The top-level directories match the examples index directly. For instance, fetch-hackathon-quickstarter/, langchain-agents/, and a2a-cart-store/ are all visible in the repository structure alongside the shared infrastructure files. ## Running Your First Example The README provides a five-step quickstart using the hackathon quickstarter example: ```bash git clone https://github.com/fetchai/innovation-lab-examples.git cd innovation-lab-examples cd fetch-hackathon-quickstarter python3 -m venv .venv && source .venv/bin/activate pip install -r requirements.txt ``` After installing dependencies, copy the environment template and edit it: ```bash cp .env.example .env ``` Then run the agent: ```bash python agents/alice/agent.py ``` The README also provides a one-command alternative using the setup script at the repository root: ```bash ./setup.sh fetch-hackathon-quickstarter ``` The README notes that Python 3.10 or higher, pip, and git are required. Some examples require API keys for services like ASI:One, OpenAI, or Stripe. Each example's .env.example documents which keys it needs, so checking that file before starting is the right first step for any example beyond the quickstarter. ## Docker Support for Any Example The repository root contains a Dockerfile and docker-compose.yml that can run any example in a container. The Dockerfile takes an EXAMPLE build argument that defaults to fetch-hackathon-quickstarter: ```dockerfile FROM python:3.11-slim ARG EXAMPLE=fetch-hackathon-quickstarter ``` The build step copies the selected example directory, installs its requirements.txt if one exists, and copies .env.example to .env if no .env is present. The docker-compose.yml passes the example name through a shell variable: ```yaml services: agent: build: context: . args: EXAMPLE: ${EXAMPLE:-fetch-hackathon-quickstarter} env_file: - ${EXAMPLE:-fetch-hackathon-quickstarter}/.env restart: unless-stopped ``` To run a different example, set the EXAMPLE variable before calling docker compose. The Dockerfile's ENTRYPOINT is python with CMD ["agent.py"], so examples that use a different entry point filename will need a command override. The Dockerfile does not document examples that use multiple agent files or that require additional system packages beyond gcc and libffi-dev. ## What the Examples Leave Out The repository is an examples collection, not a production agent framework. Nothing in the repository provides shared authentication, persistent storage, or agent orchestration that a production system would need. Each example demonstrates a pattern; adapting that pattern to a real service is the user's work. The examples index is truncated in the README, so the exact number and names of all 80+ examples are not fully visible from the top-level README alone. The full list requires browsing the directory tree or the documentation site at https://innovationlab.fetch.ai/resources/docs/intro. Payment and Web3 examples require accounts and API keys beyond a standard LLM provider key. The langchain-agents example uses Stripe, the a2a-cart-store uses Skyfire, and several examples require an Agentverse account. These dependencies mean the simplest examples install quickly, but the advanced ones have a longer onboarding path. The Dockerfile's default CMD is python agent.py, which works for single-file examples. Multi-file examples or examples with a different entry point structure will fail unless the CMD is overridden at container run time. ## Comparison with LangChain Templates, Maintenance, and License LangChain's templates repository and CrewAI's example library both offer runnable agent starter code. LangChain templates are organized around LangChain's own chain and retrieval abstractions; CrewAI examples focus on the crew orchestration model. innovation-lab-examples is organized around the uAgents framework and Fetch.ai's agent network, with LangChain and CrewAI appearing as optional integrations inside individual examples rather than as the organizing principle. A developer already committed to LangChain will find the LangChain templates collection more directly applicable; a developer exploring the Fetch.ai and Agentverse ecosystem will find innovation-lab-examples the more complete starting point. The last push to the main branch was on 2026-09-01. The repository is not archived. The repository is released under the MIT license. The LICENSE file is at the repository root. Examples that call third-party APIs (OpenAI, Anthropic, Google, Stripe) are subject to those providers' terms of service independently of the repository's own license. ## Conclusion innovation-lab-examples is the right starting point for anyone building on the Fetch.ai and uAgents ecosystem who needs a working example faster than writing one from scratch. The beginner and intermediate examples install and run with a single pip install and a .env file. Teams not using Fetch.ai or Agentverse will find the LangChain, CrewAI, and Claude examples useful but may not need the uAgents wrapping around them. Before starting an advanced example, check its .env.example for required third-party accounts, since Stripe, Skyfire, or Agentverse credentials add setup time that the difficulty label does not reflect. ## FAQ ### How do I run an innovation-lab-examples project with Docker? Set the EXAMPLE environment variable to the name of the folder you want and run docker compose up from the repository root. The docker-compose.yml passes EXAMPLE as a build argument; the Dockerfile then copies that folder and installs its dependencies. ### What API keys are required to run the examples in innovation-lab-examples? Requirements vary by example. Simple examples need only a GITHUB_TOKEN or a basic LLM API key. Advanced examples can require Stripe, Skyfire, ASI:One, or an Agentverse account. Each example documents its requirements in its own .env.example file. ### Does innovation-lab-examples include examples for Claude and Gemini? Yes. The anthropic-quickstart series covers basic Claude integration, vision, function calling, MCP, and multi-agent patterns across beginner to advanced difficulty. The gemini-quickstart series covers text, image, video, audio, and research workflows using Google Gemini. ## Sources - [fetchai/innovation-lab-examples on GitHub](https://github.com/fetchai/innovation-lab-examples) - [Issues](https://github.com/fetchai/innovation-lab-examples/issues) - [License: MIT](https://github.com/fetchai/innovation-lab-examples/blob/main/LICENSE) - [Project website](https://innovationlab.fetch.ai/resources/docs/intro) - [README](https://github.com/fetchai/innovation-lab-examples/blob/main/README.md) --- Hysen Labs editorial analysis, written from the project's own repository and release notes. Cite the canonical page: https://hysenlabs.com/projects/fetchai-innovation-lab-examples