This MCP server provides 39 tools to manage a Snipe-IT inventory system through AI-enabled interactions. It exposes assets, users, locations, licensing, reporting, and system administration endpoints, enabling your assistant to perform CRUD operations, lookups, file handling, maintenance tracking, and auditing across your Snipe-IT instance. Tools include asset management (create/read/update/delete assets, barcode/serial lookups, attachments, labels, maintenance, and license associations), inventory tracking (consumables, components, accessories with checkout/checkin flows), user and organization management (users, user assets, two-factor resets, companies, departments, groups), configuration tooling (categories, manufacturers, models, statuses, locations, suppliers, depreciations), field customization (fields and fieldsets), licensing (licenses and seats), and reporting/auditing (activity logs, status summaries, audit tracking). The MCP is designed to let an AI agent navigate the Snipe-IT API via the 39 exposed endpoints in a structured, secure way, using configurable environment variables for authentication and access control.

To use it, configure the MCP client to point at your Snipe-IT instance and provide an API token. Depending on your setup, you’ll either install the MCP server via the uvx-based GitHub install (recommended for quick starts) or run it locally from a clone using uv. Once running, your MCP client (Claude Desktop/Claude Code or Cursor) can query, create, update, or delete assets, manage users, pull reports, and perform administrative tasks through the available tools. Be sure to review SNIPEIT_ALLOWED_TOOLS to restrict exposure if needed.

Example workflows include creating a new asset, associating licenses, generating a label, exporting an activity report, or triggering a backup and system info check — all through natural-language prompts that map to the corresponding MCP tools.

Prerequisites:

• Python 3.11 or newer
• UV package manager (https://github.com/astral-sh/uv)
• Access to a Snipe-IT instance with an API token

Option A — Install directly from GitHub (recommended for quick start):

1. Install UVX and run the MCP server from GitHub: \n```bash

Install and run via UVX from GitHub (no clone required)

uvx --from git+https://github.com/jameshgordy/snipeit-mcp snipeit-mcp

2. When prompted, provide environment variables for your Snipe-IT instance.
3. Verify the server starts and is reachable by your MCP client.

Option B — Run from a local clone:
1. Clone the repository and navigate into it:
\n```bash
git clone https://github.com/jameshgordy/snipeit-mcp.git
cd snipeit-mcp

2. Install dependencies and run with UV: \n```bash uv sync uv run python server.py

3. Create a .env file (see below) or export the needed variables in your shell.

Configure environment variables (example):

```env
SNIPEIT_URL=https://your-snipeit-instance.com
SNIPEIT_TOKEN=your-api-token-here
SNIPEIT_ALLOWED_TOOLS=manage_assets,system_info  # Optional: restrict exposed tools

Optional: restrict exposed tools by editing SNIPEIT_ALLOWED_TOOLS. You can also expose all tools by leaving it unset.

Environment variables per the README are required for authentication. Common issues include incorrect SNIPEIT_URL or invalid SNIPEIT_TOKEN, which will cause API calls to fail. If tools appear missing, verify that SNIPEIT_ALLOWED_TOOLS matches the tool names in the MCP configuration and that the token has the necessary permissions. When using uvx to install directly from GitHub, ensure network access to GitHub is allowed. For local runs, ensure the working directory contains server.py if you follow the local clone path. If you need to reset access, revoke and regenerate an API token in Snipe-IT and update the environment variables accordingly. Also consider enabling backups in Snipe-IT and keeping a regular export of activity logs for auditing.