@careermax/agent-toolkit

@careermax/agent-toolkit

Connect AI agents to CareerMax -- job tracking, resume review, cover letters, interview prep, and more.

Quick Start

• Get an API key from CareerMax Settings
• Set your API key:

  export CAREERMAX_API_KEY="cmx_live_..."
  

• Run a command:

  npx @careermax/agent-toolkit jobs list
  

Setup

Getting an API Key

• Log in to CareerMax
• Go to Settings > API Keys
• Click Create API Key and copy the key (starts with cmx_live_)

Environment Variables

Variable	Required	Description
CAREERMAX_API_KEY	Yes	Your API key from CareerMax
CAREERMAX_API_URL	No	Override the API URL (defaults to CareerMax production)
CAREERMAX_LOG_LEVEL	No	Log verbosity (debug, info, warn, error)

MCP Server

The toolkit includes an MCP (Model Context Protocol) server that lets AI assistants interact with your CareerMax account directly.

Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "careermax": {
      "command": "npx",
      "args": ["-y", "@careermax/agent-toolkit", "mcp"],
      "env": {
        "CAREERMAX_API_KEY": "cmx_live_..."
      }
    }
  }
}

Config file locations:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

Cursor

Add to .cursor/mcp.json in your project root:

{
  "mcpServers": {
    "careermax": {
      "command": "npx",
      "args": ["-y", "@careermax/agent-toolkit", "mcp"],
      "env": {
        "CAREERMAX_API_KEY": "cmx_live_..."
      }
    }
  }
}

Other MCP Clients

Any MCP-compatible client can connect. Run the server with:

CAREERMAX_API_KEY="cmx_live_..." npx @careermax/agent-toolkit mcp

The server communicates over stdio using the standard MCP protocol.

CLI Usage

Setup

export CAREERMAX_API_KEY="cmx_live_..."

Or install globally:

npm install -g @careermax/agent-toolkit

Commands

Get an overview of what CareerMax can do:

careermax info

List jobs in your pipeline:

careermax jobs list
careermax jobs list --status applied
careermax jobs list --json

Review your resume with AI:

careermax resume review
careermax resume review --job-id <id>   # tailor feedback to a specific job

Generate a cover letter:

careermax cover-letter generate --job-id <id>

Create a mock interview session:

careermax interviews create --role "Software Engineer" --company Stripe
careermax interviews list

Check your credit balance:

careermax credits

Global Flags

Flag	Description
--json	Output raw JSON instead of formatted tables
--yes	Skip confirmation prompts (useful in scripts)
--api-key <key>	Override the CAREERMAX_API_KEY env var
--verbose	Enable detailed logging

Available Tools

These tools are exposed via the MCP server and can be called by any connected AI agent.

Tool	Description
careermax_info	Get an overview of CareerMax capabilities and the user's account
jobs_list	List all jobs in the user's pipeline, with optional status filter
jobs_get	Get full details for a specific job
jobs_add	Add a new job to the pipeline
jobs_update	Update a job's status, notes, or other fields
resume_review	Run an AI review of the user's resume, optionally tailored to a job
cover_letter_generate	Generate a cover letter for a specific job
interviews_create	Create a new mock interview session for a given role and company
interviews_list	List past and upcoming interview sessions
credits_get	Check the user's current credit balance and plan tier

How It Works

AI Agent (Claude, GPT, etc.)
    |
    v
MCP Server / CLI
    |  (HTTP + Bearer auth)
    v
Convex Backend (api/v1/*)
    |
    v
Same AI pipeline as the CareerMax web app

The toolkit is a thin client. All heavy lifting -- AI resume analysis, cover letter generation, interview simulation -- happens server-side on the same Convex backend that powers the CareerMax web app. Your data stays in your account.

License

MIT