CATS

You can run a FastMCP server that exposes 228 tools across 17 toolsets for the CATS API v3. It loads only the toolsets you need, keeps credentials secure, and adapts to production-grade needs with robust error handling and monitoring. This guide walks you through practical usage, installation, and key configuration details so you can start integrating CATS data efficiently.

How to use

You consume the MCP server through an MCP client by connecting to the server’s endpoint and invoking the available tools. Start by choosing the toolsets you actually need to minimize token usage and latency. For example, load only candidates, jobs, and pipelines if those cover your current workflows. You can adjust toolsets on the fly or run with all tools if you require complete API coverage.

How to install

Prerequisites: Python (3.8+), and a working internet connection to install dependencies.

# 1. Clone repository
# Replace with your actual path if needed
git clone https://example.com/cats-mcp-server.git
cd cats-mcp-server

# 2. Create and activate a virtual environment
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate

# 3. Install dependencies
pip install fastmcp httpx python-dotenv

Configuration is done via environment variables. Create a .env file and provide your CATS API details and desired toolsets.

Then you can start and test the server using the provided Python entry point. You can list toolsets, load specific toolsets, or load all tools to verify coverage.

# 1. List available toolsets
python server.py --list-toolsets

# 2. Run with default toolsets
python server.py

# 3. Run with a specific selection
python server.py --toolsets candidates,jobs,companies

# 4. Run with all toolsets
python server.py --toolsets all

Configuration and usage notes

Environment variables you may use include the base URL for the CATS API, your API key, and an optional default set of toolsets. Example values are shown below; replace them with your actual credentials.

CATS_API_BASE_URL=https://api.catsone.com/v3
CATS_API_KEY=your_cats_api_key_here

# Optional: default toolsets to load on startup
CATS_TOOLSETS=candidates,jobs,companies

If you are integrating with Claude Desktop or any environment that needs explicit MCP server configuration, provide the runtime command and environment exactly as shown in your setup. Examples below demonstrate local stdio configurations.

Security and reliability notes

Use token-based authentication and never embed credentials in code. The server is designed with asynchronous operations, structured logging, and health checks to surface issues quickly. Monitor rate limits via the X-Rate-Limit-Remaining header and implement backoffs on 429 responses to maintain smooth operation.

Troubleshooting

Common issues include missing API keys, misconfigured toolsets, or network access problems. Ensure your CATS_API_KEY is valid for API v3 access and that you are not using v2 keys. If you encounter module-not-found errors for toolsets, verify that the required toolset files exist in your deployment.

Notes

This MCP server provides comprehensive coverage for the CATS API v3 with dynamic loading. It is production-ready, supports health checks, and is designed to deploy to cloud environments with automatic GitHub-based deployments. You can adjust toolset loading to optimize token usage and performance.

Available tools

list_candidates

List candidates in the system.

get_candidate

Retrieve a single candidate by ID.

create_candidate

Create a new candidate record.

update_candidate

Update an existing candidate.

delete_candidate

Delete a candidate.

search_candidates

Search candidates using filters.

filter_candidates

Filter candidates with advanced criteria.

authorize_user

Authorize a user for access control.

list_jobs

List job postings.

get_job

Get a single job by ID.

create_job

Create a new job posting.

update_job

Update an existing job.

delete_job

Delete a job.

search_jobs

Search jobs with filters.

filter_jobs

Filter jobs with advanced criteria.

list_pipelines

List pipelines.

get_pipeline

Get a pipeline by ID.

create_pipeline

Create a pipeline.

update_pipeline

Update a pipeline.

delete_pipeline

Delete a pipeline.

list_attachments

List attachments for resources.

parse_resume

AI-powered resume parsing.

list_webhooks

List webhook subscriptions.

create_webhook

Create a webhook subscription.

get_webhook

Get a webhook by ID.

delete_webhook

Delete a webhook.

list_events

List calendar events.

create_event

Create a calendar event.