Fetch.ai’s agent cookbook: LLMs, payments, A2A, and Web3

What it does

This repository houses more than thirty self-contained example projects built on Fetch.ai’s stack—uAgents, ASI:One, Agentverse, and the A2A protocol. Each folder is a standalone pattern, ranging from a hackathon quickstarter to multi-agent systems handling Stripe payments, RAG pipelines, and Web3 integrations. It is essentially a reference architecture gallery for autonomous-agent development within the Fetch.ai ecosystem.

The interesting bit

The repo treats Fetch.ai not as a walled garden but as a hub, wiring in Anthropic, Google Gemini, OpenAI, CrewAI, LlamaIndex, and even Notte browser automation through shared uAgents patterns. That breadth turns it into an integration Rosetta Stone rather than a single toy demo.

Key highlights

• 30+ distinct examples, from beginner quickstarts to advanced multi-agent orchestration.
• Integrates major LLM providers (Anthropic, Gemini, OpenAI) alongside Fetch.ai’s own ASI:One.
• Working patterns for A2A communication, MCP servers, and payment protocols (Stripe, FET, Skyfire).
• Community contributions are explicitly welcomed via a dedicated contributors/ directory and templates.
• Docker and Docker Compose support is baked in for most examples.

Caveats

• Most examples require external API keys—ASI:One, OpenAI, Stripe, and others—so this is not a zero-config playground.
• Directory naming is inconsistent (Rag-agent, Claude Agent SDK, Crewai-agents), suggesting organic growth rather than tight editorial curation.
• The contribution guide requires starring the repository before opening a pull request, which is a mildly gamified barrier.

Verdict

Grab this if you are building on Fetch.ai and need a working reference for LLM wiring, agent-to-agent chatter, or crypto payments. Skip it if you are not invested in the Fetch.ai stack—this is ecosystem-specific glue, not generic agent theory.