Fuel Network & Sway Language MCP Server

Name: Fuel Network & Sway Language MCP Server
Author: FuelLabs

@FuelLabs

リポジトリ

Fuel Network & Sway Language MCP Server について

A Fuel MCP server which provides support for Fuel docs and various coding IDEs such as Cursor.

基本情報

カテゴリ

その他

ランタイム

node

トランスポート

stdio

公開者

FuelLabs

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "fuel-mcp-server": {
      "command": "bun",
      "args": [
        "run",
        "src/indexer.ts",
        "./docs"
      ]
    }
  }
}

ツール

ツールは検出されませんでした

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is Fuel Network & Sway Language MCP Server?

It is a Model Context Protocol server that indexes Fuel Network and Sway Language documentation into a local vector database, enabling IDE integration (e.g., Cursor) for semantic search of docs.fuel.network content directly within the development environment.

How to use Fuel Network & Sway Language MCP Server?

Clone the repository, install dependencies with bun install, index documentation with bun run src/indexer.ts, then start the MCP server with bun run src/cli.ts. Add its configuration to your MCP client (e.g., Claude or Cursor) by specifying the command bun run <absolute_path>/src/cli.ts --transport stdio.

Key features of Fuel Network & Sway Language MCP Server

Local semantic search using open‑source embeddings (Transformers.js)
No Docker required – runs solely with Bun
Fast file‑based vector storage using Vectra
Hybrid search with keyword fallback for robustness
Enhanced result filtering and formatting
Supports both STDIO and HTTP transports

Use cases of Fuel Network & Sway Language MCP Server

Searching FuelVM documentation without leaving an IDE
Retrieving Sway standard library paths and types via MCP tools
Indexing and querying local Markdown documentation offline
Integrating contextual documentation lookup into AI‑assisted coding workflows
Developing and testing Fuel‑based smart contracts with instant doc access

FAQ from Fuel Network & Sway Language MCP Server

How does it differ from online documentation search?

Unlike web‑based searches, the server indexes docs locally, works offline, and integrates directly into MCP‑compatible IDEs for faster, contextual results.

What are the runtime dependencies?

Only Bun is required. No Docker, Python, or external embedding services are needed.

Where is the documentation data stored?

Document embeddings are stored in a local vectra_index directory created after running the indexer. The path can be customized via the VECTRA_INDEX_PATH environment variable.

What transport protocols does it support?

STDIO (default) and HTTP transport (e.g., --transport http --port 3500). Health checks are available at /health.

Are there any known limits?

The server is designed for local use; it does not connect to external APIs. The search quality depends on the indexed documents and the chosen embedding model.