Building with AI - Stellar Developer Docs

Stellar provides resources to help AI assistants and Large Language Models (LLMs) understand our documentation, making it easier for you to get accurate answers about Stellar development.

Raven (MCP server)

Raven is a remote Model Context Protocol (MCP) server for AI agents, hosted at raven.stellar.buzz. It is open source at the Raven repository. This is the recommended path: connect your agent once and it gets Stellar docs plus live ecosystem data, cross-referenced into single answers on demand. Raven exposes two MCP tools:

search — find relevant Stellar docs and ecosystem information.
execute — run queries against live ecosystem data.

Under the hood it wraps Stellar Docs (via Algolia), Lumenloop, Stellar Light, and ecosystem skills, so your agent gets one cross-referenced answer instead of several disconnected sources. Unlike the static llms.txt file and the installable Stellar Skills below, Raven is a live, hosted server your agent talks to. All three are useful, and they work well together.

Try it in your browserraven.stellar.buzz/playground is a hosted chat UI where you can ask Raven questions directly in the browser, no agent setup required (sign-in required).

Connect your agent

Add the server, then sign in. Sign-in happens inside your own client (Raven uses OAuth), so the first request opens your browser to authorize. For full setup details see raven.stellar.buzz.

claude mcp add --transport http stellar-raven "https://raven.stellar.buzz/mcp"
# then run /mcp -> Authenticate -> sign in in your browser

codex mcp add stellar-raven --url "https://raven.stellar.buzz/mcp"
codex mcp login stellar-raven  # opens your browser

code --add-mcp '{"name":"stellar-raven","type":"http","url":"https://raven.stellar.buzz/mcp"}'

Add to ~/.cursor/mcp.json (or .cursor/mcp.json per project). Cursor runs the OAuth sign-in on first use.

{
  "mcpServers": {
    "stellar-raven": { "url": "https://raven.stellar.buzz/mcp" }
  }
}

Go to Settings -> Connectors -> Add custom connector, paste the endpoint, and approve the browser sign-in.

https://raven.stellar.buzz/mcp

For clients without native remote or OAuth support, bridge with mcp-remote.

{
  "mcpServers": {
    "stellar-raven": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote@latest",
        "https://raven.stellar.buzz/mcp",
        "--transport",
        "http-only"
      ]
    }
  }
}

mcp-remote

Stellar Skills

skills.stellar.org is the home for Stellar Skills—a collection of AI agent skills that give your coding assistant the right Stellar context before it writes code. They work with any AI agent, and you can browse every skill on the website with no installation required. Skills are open source and maintained in the stellar/stellar-dev-skill repository.

Official skills

The official skills cover the core areas of Stellar development:

Skill	What it covers
Soroban Smart Contracts	Writing, testing, securing, and shipping Rust smart contracts
Frontend & Wallets	Building dApps with the JavaScript SDK, wallet integration (Freighter, Stellar Wallets Kit), and smart accounts with passkeys
Stellar Assets & SAC	Classic assets, trustlines, and the Soroban Asset Contract bridge
RPC & Horizon APIs	Querying chain data via modern Stellar RPC or the legacy Horizon endpoints
Agent Payments	Monetizing AI APIs with x402 and the Machine Payments Protocol (MPP)
ZK Proofs	Verifying zero-knowledge proofs on-chain
SEPs, CAPs & Ecosystem	Standards and ecosystem integration guidance

Community skills

The ecosystem also contributes skills for popular protocols and tools. Browse the full, up-to-date list at skills.stellar.org. Community skills are maintained by their respective authors, so review each one to confirm it meets your security and maintenance requirements before using it.

stellar-build

For a more hands-on approach, the community-maintained stellar-build installer adds 46 skills plus a set of AI personas to Claude Code or Codex with a single command, guiding you through the whole journey—not just building, but also ideation and validation, all the way to launch.

curl -fsSL https://raw.githubusercontent.com/kaankacar/stellar-build/main/install.sh | bash

Installation

Skills can be installed on any AI coding tool that supports the Agent Skills standard. Each tool reads skills from its own directory:

Platform	Skills directory
Claude Code	`~/.claude/skills/`
OpenCode	`~/.config/opencode/skills/`
OpenAI Codex	`~/.codex/skills/`

Adds the skill to your tool permanently.

/plugin marketplace add stellar/stellar-dev-skill

/plugin install stellar-dev@stellar-dev

Codex reads skills from ~/.codex/skills/.

git clone https://github.com/stellar/stellar-dev-skill ~/.codex/skills/stellar-dev-skill

Works with any tool that supports the Agent Skills standard.

npx skills add https://github.com/stellar/stellar-dev-skill

Works with any tool that supports the Agent Skills standard.

npx skills add https://github.com/stellar/stellar-dev-skill

Works with any tool that supports the Agent Skills standard.

npx skills add https://github.com/stellar/stellar-dev-skill

Works with any tool that supports the Agent Skills standard.

npx skills add https://github.com/stellar/stellar-dev-skill

Manual installation

Clone the repository and copy the contents of the skills/ directory into your agent’s skills location (see table above). Once installed, your AI coding assistant will automatically have access to up-to-date Stellar development knowledge when you work on Stellar projects. For other AI coding tools like Cursor, Windsurf, or Aider, you can consider the community-maintained OpenSkills CLI, which aims to bring the Agent Skills system to a wide range of AI coding agents. OpenSkills is a third-party project and not an official Stellar offering, so be sure to review the repository and its documentation to verify that it meets your security and maintenance requirements before using it.

Using llms.txt

llms.txt is a standardized way to provide documentation context to AI systems. This file contains a structured overview of Stellar’s developer documentation optimized for LLM consumption. When you ask an AI assistant about Stellar, it can reference this file to:

Understand the structure of our documentation
Find relevant pages for your questions
Provide more accurate, up-to-date answers

Other AI tools

Any AI assistant with custom context features can benefit from Stellar’s resources. You can add llms.txt to your tool’s context settings, or paste its contents directly into conversations to give the AI knowledge about Stellar development. Popular tools like ChatGPT, Claude, Gemini, and Cursor all support adding custom context through their settings. We’re actively working on dedicated integrations for more AI tools—check back for updates or follow the Stellar Skills repository for the latest supported platforms.

Additional resources

Raven (stellar-raven) - Open source remote MCP server for Stellar docs plus live ecosystem data
Stellar Skills - Browse AI agent skills for Stellar development
stellar-dev-skill repository - Source and installation for the official skills
stellar-build - Community one-command installer with 46 skills and AI personas covering the full journey from ideation and validation to launch
Stellar Developer Discord - Ask questions and get help from the community

​Raven (MCP server)

​Connect your agent

​Stellar Skills

​Official skills

​Community skills

​stellar-build

​Installation

​Manual installation

​Using llms.txt

​Other AI tools

​Additional resources

Raven (MCP server)

Connect your agent

Stellar Skills

Official skills

Community skills

stellar-build

Installation

Manual installation

Using llms.txt

Other AI tools

Additional resources