pulsar-mcp
Enables AI assistants to control the Pulsar editor, including reading and writing files, managing selections, and interacting with project paths.
README
pulsar-mcp
Model Context Protocol server, that provides Pulsar editor tools to AI assistants.
Features
- MCP protocol: Version 2025-11-25 with tool annotations support.
- HTTP bridge: Server running inside Pulsar for direct API access.
- Standalone server: MCP server script for Claude CLI integration.
- Editor tools: Get/set content, open/save files, manage selections.
- Extensible: Other packages can register tools via
mcp-toolsservice. - Toggle tools: Enable/disable individual tools via select list. Destructive tools disabled by default.
- claude-chat integration: Auto-connects the MCP server to each Claude session via the provided service.
Installation
To install pulsar-mcp search for pulsar-mcp in the Install pane of the Pulsar settings or run ppm install pulsar-mcp. Alternatively, you can run ppm install asiloisad/pulsar-pulsar-mcp to install a package directly from the GitHub repository.
Commands
Commands available in atom-workspace:
pulsar-mcp:toggle-tools: toggle individual tools on/off,pulsar-mcp:start: start the MCP bridge server,pulsar-mcp:stop: stop the MCP bridge server,pulsar-mcp:status: show current bridge status and port.
Commands available in .pulsar-mcp:
select-list:enable-all: enable all tools,select-list:disable-all: disable all tools,select-list:reset-defaults: reset to defaults.
Built-in Tools
| Tool | Description | Default |
|---|---|---|
GetActiveEditor |
Get editor metadata (path, grammar, modified, lineCount) | Enabled |
GetOpenEditors |
Get metadata for all open text editors | Enabled |
ReadText |
Read active editor content with line pagination (use agent's file tools for other files) | Enabled |
WriteText |
Write text at cursor or replace range in active editor (use agent's file tools for other files) | Enabled |
OpenFile |
Open an existing file in editor with optional position (create=true allows new files) |
Enabled |
SaveFile |
Save a file (active editor or specific path) | Enabled |
GetSelections |
Get all selections/cursors with positions and text from active editor | Enabled |
SetSelections |
Set multiple selections/cursors at specific positions in active editor | Enabled |
CloseFile |
Close an editor tab | Disabled |
GetProjectPaths |
Get project root folders | Enabled |
AddProjectPath |
Add a folder to project roots | Enabled |
RemoveProjectPath |
Remove a folder from project roots | Disabled |
MCP Client Integration
The standalone MCP server (lib/server.js) can be used with any MCP-compatible client. The server connects to the Pulsar bridge via PULSAR_BRIDGE_PORT (default 3000). Check the actual port with pulsar-mcp:status. It auto-increments when multiple Pulsar windows are open.
Claude Code
Register the server with the Claude CLI:
claude mcp add -e PULSAR_BRIDGE_PORT=3000 pulsar -- node ~/.pulsar/packages/pulsar-mcp/lib/server.js
On Windows:
claude mcp add -e PULSAR_BRIDGE_PORT=3000 pulsar -- node "%USERPROFILE%\.pulsar\packages\pulsar-mcp\lib\server.js"
JSON config
{
"mcpServers": {
"pulsar": {
"command": "node",
"args": ["~/.pulsar/packages/pulsar-mcp/lib/server.js"],
"env": {
"PULSAR_BRIDGE_PORT": "3000"
}
}
}
}
On Windows, use "%USERPROFILE%\.pulsar\packages\pulsar-mcp\lib\server.js".
Provided Service pulsar-mcp
Provides access to the MCP bridge state: port, running status, and server script path. Used by claude-chat to auto-connect the MCP server.
In your package.json:
{
"consumedServices": {
"pulsar-mcp": {
"versions": {
"1.0.0": "consumePulsarMcp"
}
}
}
}
In your main module:
consumePulsarMcp(service) {
// Get current bridge port
const port = service.getBridgePort();
// Check if bridge is running
const running = service.isRunning();
// Get path to MCP server script
const serverPath = service.getServerPath();
}
Consumed Service mcp-tools
Other Pulsar packages can provide additional MCP tools by implementing the mcp-tools service. Each tool defines a name, description, input schema, and execute function.
In your package.json:
{
"providedServices": {
"mcp-tools": {
"versions": {
"1.0.0": "provideMcpTools"
}
}
}
}
In your main module:
module.exports = {
provideMcpTools() {
return [
{
name: "MyCustomTool",
description: "Description for the AI",
inputSchema: {
type: "object",
properties: {
param: { type: "string", description: "Parameter description" }
},
required: ["param"]
},
annotations: { readOnlyHint: true },
execute({ param }) {
// Tool implementation
return { result: "data" };
}
}
];
}
}
Contributing
Got ideas to make this package better, found a bug, or want to help add new features? Just drop your thoughts on GitHub. Any feedback is welcome!
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.