Career-Agents

The Open-Source Career Operating System (Career OS)

🚀 Overview

Career-Agents is an open-source Career Operating System (Career OS) designed to modularize, audit, and automate professional growth tasks. It replaces unstructured prompt structures and static text templates with a unified CLI utility, an stdio-based Model Context Protocol (MCP) server, and a premium Next.js web dashboard.

The system parses resumes for ATS compatibility, audits public GitHub repositories, analyzes LinkedIn profile copy, and runs mock STAR interviews. By synchronizing all metrics into a local profile state, it calculates an overall Career Readiness Score and generates custom actionable roadmaps.

🎨 Visual Interface & Modules

The platform is divided into core feature modules that work seamlessly together:

1. Resume Studio

• ATS Reviewer: Analyzes formatting blockers (multi-column, tables, icons) that trip up parsing systems.
• Weak Bullet Auditor: Identifies weak action verbs and missing metrics, suggesting instant STAR (Situation, Task, Action, Result) revisions.
• FAANG Calibrator: Validates signaling against target company competencies.

2. GitHub Analyzer

• Portfolio Scorer: Computes language diversity, contribution maps, and repository description completeness.
• README Auditor: Scores documentation visibility and project packaging benchmarks.

3. LinkedIn Optimizer

• Tagline Generator: Provides professional pipe-separated headline formats.
• Visibility Auditor: Computes search indexing density and flag patterns.

4. Mock Interview Lab

• STAR Simulator: Conducts mock interviews in the CLI or Web Dashboard using the STAR framework.
• Integrated Code Board: Code editing panel for technical and system design rounds.
• Scorecard Evaluation: AI evaluates response content and communication quality.

💻 Visual Product Preview

For a demonstration of features, guides, and layouts in action:

• Interactive Dashboard: Refer to ./docs/images/dashboard_preview.gif.
• Resume Studio: Refer to ./docs/images/resume_studio_preview.gif.
• GitHub Audit: Refer to ./docs/images/github_wrapped_preview.gif.
• LinkedIn Optimizer: Refer to ./docs/images/linkedin_optimizer_preview.gif.
• Mock Interview Lab: Refer to ./docs/images/mock_interview_preview.gif.

🛠️ CLI Reference

1. General Diagnostic and Info

• List: career-agents list — Lists registered divisions, workflows, and available agent coaches.
• Doctor: career-agents doctor — Verifies Node.js path settings, environment config, and dependencies.
• Recommend: career-agents recommend --skills "react,node" — Recommends career coaches and tracks.

2. Resume Review

• Review: career-agents review resume.pdf — Run full ATS and formatting scanner.
• Score: career-agents score resume.pdf — Calculates ATS compatibility score (0-100%).
• Improve: career-agents improve resume.pdf — Generates suggested bullet rewrites using active verbs.

3. Portfolio & Profile Audits

• GitHub: career-agents github <username> — Grades repository documentation, language diversity, and stars.
• LinkedIn: career-agents linkedin profile-copy.txt — Analyzes headline positioning and summary copy.

4. Roadmaps & Interviews

• Mock: career-agents mock stripe technical — Starts interactive terminal mock interview drill.
• Roadmap: career-agents roadmap google — Outputs target study roadmap checkoff list.

🌐 Next.js Web Dashboard

The repository includes a modern Next.js 14 Web Dashboard located under apps/web/. It connects the modular engines into a premium single-page application.

Features

• Zustand Store: Persistent local state management saving metrics, applications, and logs locally.
• Local-First Persistence: High-speed JSON-based storage wrapper mirroring Prisma client models.
• AI Provider Selector: Support for 9 LLM backends (Groq, OpenAI, Anthropic, Gemini, DeepSeek, etc.).
• Command Palette: Press Cmd+K / Ctrl+K to search pages, trigger audits, and switch views.

Local Development Setup

1. Navigate to the web app:
   hljs language-bash

   Copy

   cd apps/web
   

2. Install dependencies:
   hljs language-bash

   Copy

   npm install
   

3. Configure environment:
   hljs language-bash

   Copy

   cp .env.example .env
   # Add your API keys and database parameters
   

4. Launch development server:
   hljs language-bash

   Copy

   npm run dev
   

   Open http://localhost:3000 in your browser.

🔌 Model Context Protocol (MCP)

Expose Career OS tools and resources directly to LLM clients (like Claude Desktop or Cursor).

Claude Desktop Setup

Add the config block inside %APPDATA%/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "career-agents": {
      "command": "npx",
      "args": ["-y", "career-agents", "mcp"]
    }
  }
}

Cursor Setup

1. Navigate to Settings → Features → MCP.
2. Click Add New MCP Server.
3. Fill details:
   • Name: career-agents
   • Type: stdio
   • Command: npx -y career-agents mcp

📐 System Architecture

The project is structured modularly to separate execution runtimes, UI frameworks, and templates:

graph TD
    CLI[scripts/cli.js] --> Core[packages/core/]
    CLI --> Resume[packages/resume/]
    CLI --> GitHub[packages/github/]
    CLI --> LinkedIn[packages/linkedin/]
    CLI --> Interview[packages/interview/]
    CLI --> Dashboard[packages/dashboard/]
    CLI --> Reports[packages/reports/]
    CLI --> Plugins[packages/plugins/]
    CLI --> Telemetry[packages/telemetry/]
    CLI --> MCP[packages/mcp/]
    
    MCP --> Resume
    MCP --> GitHub
    MCP --> LinkedIn
    MCP --> Interview
    MCP --> Dashboard

    WebApp[apps/web/] --> API[app/api/]
    API --> AI[lib/ai.ts]
    AI --> ExternalAPIs[Groq/Gemini/OpenAI]

📚 Reference Guides

Comprehensive Technical Manuals (no emojis):

• System Architecture Guide: Design patterns, directory maps, and system flows.
• CLI Reference Manual: Command catalogs, options, and parameters.
• MCP Integration Guide: Stdio configuration for Cursor, Windsurf, and VS Code.
• Developer Plugin Guide: Custom plugin scans, hook registrations, and schemas.
• Dashboard Guide: State synchronization detail.
• Resume Studio Guide: ATS scoring rules.
• GitHub Analyzer Guide: Repository grading metrics.
• LinkedIn Optimizer Guide: Search visibility index.
• Interview Lab Guide: Readline simulation loop.
• Job Tracker Guide: Applications status tracking.
• FAQ Reference: Frequently asked questions.
• Troubleshooting Guide: Diagnostics and path fixes.

👥 Contributing

We welcome contributions! Please review the Contributor Guide and the AI Contributor Rules before submitting Pull Requests.

📄 License

This repository is licensed under the MIT License — see LICENSE for details.