Developer Tools

Trackly CLI + MCP Server

Search jobs, track applications, and query your job hunt from the terminal or your AI agent.

Install

npm install -g trackly-cli

Quick Start

# Authenticate
trackly login

# Browse recent jobs
trackly jobs

# AI-powered natural language search
trackly ask "PM jobs at fintech companies in SF"

Use an API Key Instead

# Save an API key locally
trackly config --api-key trk_xxxxxxxxxxxxxxxxxxxx

# Or use one just for this command
TRACKLY_API_KEY=trk_xxxxxxxxxxxxxxxxxxxx trackly jobs --json

Commands

CommandDescription
trackly loginAuthenticate with your Trackly account
trackly jobsList recent job postings with filters
trackly job <id>View details for a specific job
trackly jobs <id>Alias for viewing a specific job
trackly companiesList monitored companies
trackly companies search <q>Semantic company search
trackly search <query>Alias for semantic company search
trackly ask <question>Natural language job search (AI-powered)
trackly statsView your job tracking metrics
trackly statusAlias for stats
trackly save <id>Save a job to your tracker
trackly apply <id>Mark a job as applied
trackly dismiss <id>Dismiss a job
trackly api-key createGenerate an API key
trackly config --api-key trk_xxxStore an API key for future commands
trackly whoamiShow the current authenticated user
trackly versionShow the installed CLI version
trackly logoutSign out and clear credentials

MCP Server Setup

The Trackly CLI includes a built-in MCP (Model Context Protocol) server. This lets Claude Code, Cursor, and other MCP-compatible clients query your job data directly.

Start it manually with trackly mcp, or add one of the configs below.

~/.claude/settings.json
{
  "mcpServers": {
    "trackly": {
      "command": "trackly",
      "args": ["mcp"]
    }
  }
}
.cursor/mcp.json
{
  "mcpServers": {
    "trackly": {
      "command": "trackly",
      "args": ["mcp"]
    }
  }
}

Agent Integration

Once the MCP server is configured, Claude Code and Cursor can access your Trackly data natively. Ask questions in natural language and your agent will use the Trackly tools to fetch real-time job data.

Example prompts

"Find me product manager roles at Series B+ companies in New York"

"What new engineering jobs were posted today?"

"Show my application stats and pipeline"

Requirements