blueiris-mcp
MCP server for Blue Iris NVR, enabling camera health checks, live/historical snapshots, clip/alert search, and PTZ control via the Blue Iris JSON API.
README
blueiris-mcp
An MCP server for Blue Iris, the Windows NVR/camera-management software. Lets an LLM check camera health, pull live or historical snapshots, search the clip/alert archive, and drive PTZ cameras — all through Blue Iris's existing JSON API, no extra software on the Blue Iris box required.
Blue Iris's JSON API is undocumented enough that its session-handshake auth scheme (a custom two-step MD5 challenge, not a bearer token or HTTP digest) has to be reverse-engineered by reading network traffic. This package does that once so you don't have to.
Why
Blue Iris already has a web UI and a JSON API, but neither is something an LLM agent can use directly. This wraps the API as MCP tools so an agent can answer questions like:
- "Is the front door camera actually seeing anything right now, or is it stuck on the no-signal placeholder?"
- "Show me what the driveway camera saw at 6pm yesterday."
- "Who showed up at the front door today?" (Blue Iris's own AI recognition results are in the alert log's
memofield, e.g."John:74%") - "Pan the workshop camera left for a second and grab a frame."
Tools
| Tool | Description |
|---|---|
list_cameras |
Configured cameras with live health: is_no_signal (the actual "no video" placeholder state, distinct from is_online), fps, error, ptz_capable. |
get_snapshot |
Live or historical JPEG frame (pos_ms = Unix epoch ms for historical). Returned as image content, not a file path. |
list_clips |
Recorded segments for a camera, newest first, across Blue Iris's full retention window for that camera. |
list_alerts |
Motion/AI alerts for a camera, newest first, including Blue Iris's own recognition memo when present. |
ptz |
Directional nudge, home, zoom in/out, stop, or jump to a saved preset. |
Setup
Requires a Blue Iris account with JSON API access (the same username/password as the web UI login).
pip install blueiris-mcp
# or: uvx blueiris-mcp
Configure via environment variables:
| Variable | Required | Default |
|---|---|---|
BI_URL |
no | http://localhost:81 |
BI_USER |
yes | — |
BI_PASSWORD |
yes | — |
Claude Code / Claude Desktop (~/.claude.json or claude_desktop_config.json)
{
"mcpServers": {
"blueiris": {
"command": "uvx",
"args": ["blueiris-mcp"],
"env": {
"BI_URL": "http://192.168.1.50:81",
"BI_USER": "your-username",
"BI_PASSWORD": "your-password"
}
}
}
}
Hermes (~/.hermes/config.yaml)
Hermes' mcp_servers config supports the same stdio transport:
mcp_servers:
blueiris:
command: uvx
args: ["blueiris-mcp"]
env:
BI_URL: "http://192.168.1.50:81"
BI_USER: "your-username"
BI_PASSWORD: "your-password"
Notes on the Blue Iris API
- Auth:
POST /json {"cmd":"login"}returns asessiontoken. ThenPOST /json {"cmd":"login","session":<id>,"response":<hash>}wherehash = md5(f"{user}:{session}:{password}")— a single MD5 pass, no realm string. This client re-authenticates on every call rather than caching a session, since call volume through an MCP tool is inherently low. list_clips/list_alerts(Blue Iris'scliplist/alertlistcommands) have no date-range parameters — they return everything currently retained on disk for that camera. Retention is whatever Blue Iris is configured to keep (commonly a few days to a couple of weeks, storage-dependent).- If your setup has another process also driving a camera's PTZ (e.g. an automated monitoring script), avoid sending PTZ commands to that camera at the same time from here — Blue Iris does not arbitrate between simultaneous PTZ sources, and the result is erratic movement.
Development
uv venv && uv pip install -e ".[dev]"
uv run pytest
Tests are fully mocked (via respx) — no live Blue Iris instance needed.
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.