sid-mnnit/mcp

Developer tools

이 페이지는 AI로 자동 번역되었습니다. 기술적인 세부 사항은 원문을 확인하세요. 원문 보기

0 stars 2 forks 품질 58 트렌드 20

Repository Install

We are building a MCP tool .

개요

A sophisticated, multi-agent AI platform that intelligently manages your career by orchestrating job searches, email communications, calendar events, and personalized insights. Built with modern AI/LLM integrations, MCP (Model Context Protocol) servers, and a full-stack architecture. is an intelligent career management system that combines: - - Finds and ranks job opportunities based on your profile - - Drafts, manages, and prioritizes job-related communications - - Automatically extracts and schedules interview events from emails - - Ask Jobsy chatbot for career guidance and insights - - Intelligent document processing and profile management - - Multi-agent coordination for complex career workflows The platform uses for secure, standardized communication with external services (Gmail, Google Calendar, job boards) and leverages multiple LLM providers for intelligent decision-making.

README

원문 README의 일부입니다. 원문 보기.

Jobsy 3.0

Job hunting is mostly admin: checking inboxes for recruiter replies, keeping track of which applications went where, scheduling interviews, tweaking your resume for the tenth time. Jobsy is my attempt to take that admin off your plate. You connect your Gmail and Google Calendar, upload your resume, and then just ask it things in plain English, like “did Amazon get back to me?”, “when is my next interview?” or “find me backend roles in Bangalore that match my skills”.

I built it as a full-stack project to learn how to wire an LLM to real tools (Gmail, Calendar, a job search API, a resume parser) using the Model Context Protocol (MCP).

What it can do

Ask Jobsy: a chat assistant that understands what you want and runs the right action.
Email: syncs job-related emails, summarises them into a digest, answers questions about a specific company’s email, and drafts replies. Nothing is sent until you approve it.
Calendar: pulls interviews out of emails, creates events, and shows what’s coming up.
Jobs: search by keyword and location, get recommendations based on your profile, and see jobs ranked against your skills.
Resume: parses an uploaded resume, gives an ATS score, and suggests improvements.
Interview prep: pulls together your resume and the relevant emails to prepare you for an upcoming interview.
Application tracking: stats and follow-up nudges for the applications you’ve made.

How it’s put together

There are three services plus the MCP servers they talk to:

React (Vite)  ──►  Node/Express backend  ──►  Chat API (FastAPI)
  frontend          auth, profile, MongoDB       memory
                    Google OAuth, email sync          │
                           ▲                          ▼
                           └────────────  Orchestrator (FastAPI)
                                            pipelines for email, jobs,
                                            calendar, resume
                                                      │
                                      MCP servers (Gmail, Job Search,
                                      Resume, Calendar)

Folder	What lives there
`frontend/`	React 19 + Vite app (dashboard, jobs, emails, resume, settings)
`backend/`	Node/Express API with MongoDB, Google OAuth, Cloudinary uploads and the email sync queue
`client/ask_jobsy/`	The chat API and Redis-backed conversation memory
`client/orchestrator/`	The ReAct agent and the pipelines that do the real work (email, jobs, calendar, resume)
`client/llm/`	OpenAI and Groq clients, prompts, job matching and ranking
`mcp_servers/`	`gmail_mcp`, `job_search_mcp`, `resume_mcp`, `calendar_mcp`
`docs/`	Notes on architecture, agent flow and MCP contracts

How a chat message is handled

The message goes to the chat API, along with recent history from Redis.

GitHub에서 전체 README 보기

설치

This server does not publish a one-line install command.

Open the repository installation guide MCP 초보자라면? mcpskill.org 가이드부터 시작하세요 →