# Senzii Scheduling [Health: Active]

**Category:** 💻 Developer Tools  
**Repository:** https://github.com/Senzii-App/app  
**GitHub Stars:** 2  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/senzii-scheduling-2

## Description
Open-source staff scheduling: skills matching, shifts, certifications. Self-hostable, MCP-native.

## Claude Desktop Quick Installation
No confirmed install details yet, so no config is published. Follow the setup steps in the repository above.

## Documentation

## What Senzii Scheduling does

Senzii Scheduling MCP server gives an AI agent an operational interface for staff scheduling. The MCP server is included in the Senzii Scheduling application repository and uses the same PostgreSQL schema as the web application. It is designed for scheduling workflows rather than unrestricted database access.

The available tools cover staff management, candidate selection, shifts, assignments, clients, staffing requests, certifications, and work sites. Candidate searches can rank staff using tags, certifications, proximity, and availability. The platform also supports organization metrics and a restricted SQL tool for platform operators or super users.

## How it works

The application uses FastAPI and PostgreSQL through asyncpg. Sessions are cookie-based and stored in PostgreSQL. The MCP implementation uses the Python MCP SDK and exposes a Streamable HTTP interface.

Each tool query is scoped to the authenticated organization. The organization identifier is added on the server side, preventing a tool call from accessing another organization’s data. Assignment creation intentionally produces pending assignments. An agent can propose staffing decisions, but a person must confirm them before those assignments become part of the live schedule.

The matching workflow can combine staff tags and certifications with location proximity and availability. Staffing requests can be created, converted, or cancelled, while client and staff magic-link messages can be sent when the optional email capability is configured.

## Setup and configuration

PostgreSQL and a session secret are required. The repository’s quick-start process creates a Python virtual environment, installs the requirements file, and copies `.env.example` to `.env`. Set `DATABASE_URL` and `SESSION_SECRET`, then start the FastAPI application with Uvicorn. Start the MCP process separately with `python -m mcp_server.main`.

Resend is optional and only needed for email notifications. Geocoding uses the public OpenStreetMap Nominatim endpoint by default and does not require an API key; its usage policy calls for a User-Agent. Deployment scripts support Azure and Linode targets.

## Tools and capabilities

Senzii Scheduling MCP server provides 28 tools, including:

- Staff listing, retrieval, creation, deletion, availability updates, and certification assignment
- Candidate ranking for shifts
- Shift listing, assignment creation, unassignment, and deletion
- Client and staffing-request management
- Certification and work-site management
- Organization metrics
- Select-only SQL execution for authorized platform operators

Tool definitions are stored in `mcp_server/tools/tools.toml`, with handlers under `mcp_server/tools/`, so maintainers can extend the tool set within the repository’s architecture.

## Limitations and notes

The MCP server requires PostgreSQL. Email-related functions depend on an optional Resend API key, and Nominatim usage must follow the public endpoint’s policy. Assignment tools do not directly publish confirmed schedules: new assignments are pending until a human confirms them. The SQL tool is restricted to platform operators or super users and is SELECT-only.

Senzii Scheduling MCP server should be evaluated as part of the full Senzii application because it shares the application database and authentication model. The README identifies Claude, ChatGPT, and other MCP-compatible clients, but does not provide client-specific setup instructions.

_Full upstream README: https://allmcps.com/mcp/senzii-scheduling-2/readme_