Skip to content
| Marketplace
Sign in
Visual Studio Code>Other>@chrom-ar/sparkNew to Visual Studio Code? Get it now.
@chrom-ar/spark

@chrom-ar/spark

Chromar

|
2 installs
| (0) | Free
Chromar turbocharges web3 development with AI-powered tools, vivid docs, transaction simulation, security analysis, and gas optimization for a faster, smarter coding experience.
Installation
Launch VS Code Quick Open (Ctrl+P), paste the following command, and press enter.
Copied to clipboard
More Info

Chromar Spark

A VS Code extension that turbo-charges Web3 development with AI-powered chat, contextual documentation hovers and seamless code generation.

Features

  1. AI Chat Assistant (spark) • Available from the VS Code Chat view as a participant named spark. • Detects if the question is Web3-related. • For Web3 questions it performs a Retrieval-Augmented Generation (RAG) query against your Model-Context-Protocol (MCP) server, then streams the answer (or generated code) directly into the active editor. • For general programming questions it answers with standard LLM responses.

  2. Documentation on Hover • Hover over a line of JavaScript, TypeScript or Solidity to receive a concise explanation and links to relevant docs fetched from the MCP server.

  3. Language Support • JavaScript • TypeScript • Solidity

Development requirements

  • VS Code ≥ 1.102.0
  • Node.js ≥ 18 (for local development / building)
  • Access to a running MCP server that provides the tools get_available_sources and perform_rag_query.

Installation

# clone
git clone https://github.com/chrom-ar/spark.git
cd spark

npm install
npm run compile

# launch the extension development host
code .

To install the extension system-wide, run vsce package then install the generated .vsix file from VS Code (⇧⌘P → Extensions: Install from VSIX…).

Configuration

Setting Description Default
spark.serverUrl URL of the MCP Server SSE endpoint. https://mcp-docs.chrom.ar/sse
  1. Open Settings → search for Spark.
  2. Enter the URL of your MCP server.
  3. Reload the window.

Using Spark

Chat

  1. Open the Chat view in VS Code.
  2. Select spark as the participant.
  3. Ask a question or paste a code snippet.
  4. For Web3 questions, Spark injects any referenced files & docs and can stream updated code straight into the current editor.

Hover Docs

Simply hover your cursor over a line of JS/TS/Solidity code to see documentation.

Development

Common tasks:

Command Description
npm run watch Incremental compilation while you work
npm run lint / npm run lint:fix Lint source files
npm test Run unit & integration tests (uses the VS Code Test CLI)

Feel free to open issues or pull requests to improve Spark!

  • Contact us
  • Jobs
  • Privacy
  • Manage cookies
  • Terms of use
  • Trademarks
© 2025 Microsoft