SDL

v1.0.0Coding Agentsstable

Symbol Delta Ledger (SDL-MCP) gives coding agents the right code context, not your entire repo. It turns sprawling codebases into compact, high-signal context that saves tokens, speeds up workflows, and improves agent output.

agent-contextagent-toolsagentic-codingagentic-engineeringagentic-workflow
Share:
305
Stars
0
Downloads
0
Weekly
0/5

What is SDL?

SDL is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to symbol delta ledger (sdl-mcp) gives coding agents the right code context, not your entire repo. it turns sprawling codebases into compact, high-signal context that saves tokens, speeds up workflows, a...

Symbol Delta Ledger (SDL-MCP) gives coding agents the right code context, not your entire repo. It turns sprawling codebases into compact, high-signal context that saves tokens, speeds up workflows, and improves agent output.

This server falls under the Coding Agents category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • Symbol Delta Ledger (SDL-MCP) gives coding agents the right

Use Cases

Extract minimal code context for agents.
Reduce token usage in codebases.
Provide high-signal code information.
GlitterKill

Maintainer

LicenseNOASSERTION
Languagetypescript
Versionv1.0.0
UpdatedMay 21, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx sdl

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use SDL

Symbol Delta Ledger (SDL-MCP) is a TypeScript MCP server that gives coding agents precisely the code context they need — no more, no less — by indexing a repository's symbol graph and serving metadata cards, dependency slices, and code excerpts on demand instead of dumping entire files into context. It exposes 38+ tools organized around symbol search, context slicing, repository management, live buffer integration, sandboxed runtime execution across 16 languages, semantic diff analysis, and cross-session memory. Teams use it to reduce token consumption by up to 20x, speed up agentic coding workflows, and give AI agents a structured view of large codebases without exceeding context limits.

Prerequisites

  • Node.js 24 or higher installed
  • npm or npx available on your PATH
  • A code repository to index (supports TypeScript, JavaScript, Python, and other languages)
  • An MCP-compatible AI client such as Claude Desktop, Cursor, or VS Code with MCP support
1

Install sdl-mcp globally

Install the sdl-mcp package globally using npm. Node.js 24+ is required. Ignore any tree-sitter peer dependency warnings — they are harmless.

npm install -g sdl-mcp
2

Initialize and index your repository

Run the init command in your project root. The -y flag accepts defaults and --auto-index automatically detects your languages and starts indexing the repository's symbol graph.

cd /path/to/your/project
sdl-mcp init -y --auto-index
3

Start the MCP server

Launch the SDL MCP server in stdio mode so your AI client can connect to it. This command is what your MCP client configuration will invoke.

sdl-mcp serve --stdio
4

Add sdl-mcp to your AI client configuration

Open your AI client's MCP configuration file and add an entry for sdl-mcp. The server should be pointed at your project directory where the .sdl/config.json was generated by the init step.

{
  "mcpServers": {
    "sdl": {
      "command": "sdl-mcp",
      "args": ["serve", "--stdio"],
      "cwd": "/path/to/your/project"
    }
  }
}
5

Verify indexing and review the config (optional)

Check the generated .sdl/config.json to review detected languages and indexing settings. You can adjust excludePatterns, enable memory, or switch to gateway mode for a more compact 6-tool interface.

SDL Examples

Client configuration

Claude Desktop configuration for SDL-MCP running in a specific project directory. The cwd ensures the server finds the .sdl/config.json created during initialization.

{
  "mcpServers": {
    "sdl": {
      "command": "sdl-mcp",
      "args": ["serve", "--stdio"],
      "cwd": "/path/to/your/project",
      "env": {}
    }
  }
}

Prompts to try

After connecting, use these prompts to leverage SDL-MCP's symbol-level code intelligence in your AI client.

- "What does the validateToken function do? Find it and give me a summary"
- "Show me everything that calls the authenticateUser function"
- "I need to fix the auth middleware — give me the relevant code context with dependencies"
- "Analyze the blast radius of changing the UserRepository.findById method"
- "Run the tests for the authentication module and summarize the results"
- "What changed semantically in the last git diff for this repository?"

Troubleshooting SDL

Node.js version error on startup

SDL-MCP requires Node.js 24 or higher. Run 'node --version' to check your current version. Use nvm to install and switch to Node.js 24: 'nvm install 24 && nvm use 24', then reinstall sdl-mcp.

Symbols not found after indexing

Re-run 'sdl-mcp init -y --auto-index' from the project root directory. Check .sdl/config.json to confirm the correct languages are detected and that node_modules and dist are in the excludePatterns list. Large repos may take several minutes to index.

Tree-sitter peer dependency warnings during install

These warnings are harmless and do not affect functionality. If the install fails entirely, add the --legacy-peer-deps flag: 'npm install -g sdl-mcp --legacy-peer-deps'.

Frequently Asked Questions about SDL

What is SDL?

SDL is a Model Context Protocol (MCP) server that symbol delta ledger (sdl-mcp) gives coding agents the right code context, not your entire repo. it turns sprawling codebases into compact, high-signal context that saves tokens, speeds up workflows, and improves agent output. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install SDL?

Follow the installation instructions on the SDL GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with SDL?

SDL works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is SDL free to use?

Yes, SDL is open source and available under the NOASSERTION license. You can use it freely in both personal and commercial projects.

Browse More Coding Agents MCP Servers

Explore all coding agents servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "sdl": { "command": "npx", "args": ["-y", "sdl"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use SDL?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides