Hackathon MCP server for hh.ru vacancy search
SecondHackaton-HackAI
A small MCP server and agent from a Cloud.ru hackathon: hh.ru vacancy search and details through the open API, no access token needed
Low risk
We rate an entry low when it mostly gives the agent instructions and reference material.
Why this level
- Only reads public hh.ru vacancies, no applications or writes
- Works with no hh.ru account access token
Install
Manual install
pip install -r requirements.txtInstalling dependencies from the README.
This is third-party code. Review the repository files before installing.
What it does
The project was built at the Cloud.ru AI Devtools Hack and consists of an MCP server (main.py) plus a separate test AI agent (app/main_agent_test.py) on top of it. The app/tools folder structure shows search_vacancies, search_vacancies_advanced, get_vacancy_details and list_uploads tools. The README calls the hh.ru API "open to use," meaning the server hits public vacancy search with no token of its own, while running the LLM agent only needs an LLM_API_KEY in .env. Documentation is limited to a README with run commands; there's no detailed description of tool parameters in the repository.
Who it is for. For people who want a minimal working hh.ru vacancy-search MCP server as a base for their own project, not a finished polished solution.
Good fit when
- You want a simple example hh.ru vacancy-search MCP server with no token, as a starting point
- You want a ready Docker Compose setup and both stdio and SSE transports as a base
- You're fine with minimal documentation and the project's hackathon origin
Not a fit when
- You need detailed documentation of tools and parameters: the repository only has a README with run commands
- You need an actively maintained project: the last push is from December 2025, right after the hackathon
- You need advanced features like applying, resumes or analytics: the toolset is limited to search and vacancy details
Example request
Find backend developer vacancies on hh.ru and show details for the first threeLimitations
Documentation is limited to a short README; tool parameters need to be checked directly in the app/tools code. The project hasn't been updated since December 2025, right after the Cloud.ru hackathon. A .env file is committed to the repository (without a real key), which is worth checking before using someone else's fork.
How to disable. Stop the main.py process and remove the server from your MCP client configuration.
MCP
- Transport
- stdio, sse
- Authentication
- not required
| Environment variables | |
|---|---|
| LLM_API_KEY secret | An LLM provider key for the test agent, not needed by the MCP server itself. |
Security check
- Only reads public hh.ru vacancies, no applications or writes
- Works with no hh.ru account access token
README in short
The short Russian README describes the project as an agent for automatic hh.ru vacancy search from the Cloud.ru AI Devtools Hack, names the team, and gives commands for installing dependencies, setting LLM_API_KEY, and running the test agent and MCP server separately. It links to a video install walkthrough. MIT license.
FAQ
Do I need an hh.ru token?
No, the README calls the hh.ru API open to use, and the server searches vacancies with no token of its own.
What is LLM_API_KEY for?
It's only needed for the test agent in app/main_agent_test.py, not for the MCP server itself.
Related
A self-hosted knowledge base with block-level references and a built-in MCP server for connecting AI agents to your notes
A CLI for every Google Workspace API with JSON output and agent skills: Drive, Gmail, Calendar, Sheets and more
Local search over Markdown notes, docs and meeting transcripts: keywords, semantic search and reranking, with an MCP server
A task manager for AI-driven development: breaks a PRD into dependent tasks and guides the agent through them via MCP or CLI