karthik.dev
Back to projects
Open Source

Career Agents

Open-source AI Career OS — 167 agents, 19 divisions, MCP server, Spoken Voice Lab, global CLI on NPM

GitHub
TypeScriptPythonNext.jsOpenAI APIMCP (Model Context Protocol)Node.jsPostgreSQLRedisDockerNPM CLI

Overview

Career-Agents is an open-source AI Career Operating System and global CLI (npm install -g career-agents) that empowers students, developers, and professionals to optimize resumes, prepare for interviews, analyze job compatibility, and navigate career paths using 167+ specialized AI agents across 19 functional divisions. Includes a Spoken Voice Lab (27 languages), ATS Resume Studio, FAANG Interview Tracks, GitHub Portfolio Auditor, Career Knowledge Graph, and a production MCP server compatible with Claude Desktop, Cursor, and Windsurf.

Problem

Job seekers juggle disconnected tools for resume writing, interview prep, LinkedIn optimization, and tracking applications — redoing the same context-gathering work for every tool. No single AI-native platform coordinates these tasks with specialist agents that share candidate context.

Solution

A pipeline of 167+ cooperating AI agents across 19 divisions — Resume Agent, Interview Agent, LinkedIn Agent, Voice Lab, Job Tracker, and more — sharing a single candidate context object, each specializing in one transformation. Backed by a production MCP server exposing 10 tools and a global NPM CLI for instant access.

Architecture

  1. 1Global CLI (npm install -g career-agents) — entry point for all agent commands
  2. 2MCP Server exposing 10 tools for Claude Desktop, Cursor, and Windsurf integration
  3. 3167+ specialized AI agents across 19 divisions with typed JSON contract handoffs
  4. 4Career Orchestrator decomposes the request and dispatches to specialist agents
  5. 5ATS Resume Studio — keyword scoring, bullet optimization, format analysis
  6. 6FAANG Interview Tracks — role-specific question banks + AI-generated model answers
  7. 7Spoken Voice Lab — 27-language voice interaction layer for agent commands
  8. 8Career Knowledge Graph — entity-linked career intelligence and path recommendations
  9. 9GitHub Portfolio Auditor — AI-powered repository and profile analysis
  10. 10PostgreSQL + Redis backing agent state, RBAC, and analytics

Technical Challenges

  • Keeping 167 agent handoffs deterministic — typed JSON contracts prevent context drift across division boundaries.
  • Designing an MCP server that works reliably across Claude Desktop, Cursor, and Windsurf with consistent tool schemas.
  • Controlling latency and cost across multi-agent chains by batching and caching intermediate outputs.
  • Building a 27-language voice interface with consistent agent persona across languages.

Results

  • 39 GitHub stars, 7 forks — top open-source AI career platform.
  • Published as a global CLI on NPM — instantly accessible with npm install -g career-agents.
  • Production MCP server integrates with Claude Desktop, Cursor, and Windsurf.
  • 958 commits in August 2026 alone — actively maintained flagship project.
  • Career-Agents commands: doctor, assess, recommend, company, path — full career OS from the terminal.