ipcam-mcp
MCP server for local IP cameras (ONVIF/RTSP): snapshots, recordings, pan/tilt, presets
Install / Use
claude mcp add aidev-iprokszin -- npx -y github:aidev-iprokszin/ipcam-mcpIf the server publishes to npm under a different name, use that package instead — check the repo README.
MCP Server
Model Context Protocol server
Quality Score
Category
Development & EngineeringSupported Platforms
Tags
Our assessment of ipcam-mcp
ipcam-mcp scores 80/100 on our quality scale, 3557th of 4,698 Development & Engineering skills we index.
Its MCP Server is 22 KB long, well organised into 21 sections with 5 code examples: a thorough specification that gives an agent plenty to work with.
It has 3 GitHub stars, so there is little community track record yet; judge it on its content.
Maintenance, license and trust
- The repository was last updated 6 days ago, so ipcam-mcp is actively maintained.
- Our last check on 2026-11-12 found the source still online.
- It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
- Its trust signals score 87/100, with 2 cautions from licensing, adoption, age or documentation. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.
Safety scan
No issues foundOur scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.
AI review by kimi-k2.7-code on 2026-11-10. Automated pattern scan on 2026-11-09. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.
ipcam-mcp compared with similar skills
All 4 of these similar skills score higher than ipcam-mcp; compare them before choosing.
| Skill | Score | Stars | Updated | Format |
|---|---|---|---|---|
| ipcam-mcp (this skill)by aidev-iprokszin | 80 | 3 | 6d ago | MCP Server |
| taste-skillby Leonxlnx | 100 | 116.8k | 8d ago | CLAUDE.md |
| Agent-Reachby Panniantong | 100 | 112.3k | 8d ago | CLAUDE.md |
| rufloby ruvnet | 100 | 81.9k | today | CLAUDE.md |
| headroomby headroomlabs-ai | 100 | 79.8k | today | CLAUDE.md |
Frequently asked questions
- How do I install ipcam-mcp?
- Run
claude mcp add aidev-iprokszin -- npx -y github:aidev-iprokszin/ipcam-mcp. The install tabs above show the steps for each supported agent. - Which AI agents does ipcam-mcp work with?
- It is written for Claude Code and Claude Desktop, as a MCP Server file. Other agents that read the same format can often use it too.
- Is ipcam-mcp safe to use?
- Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It is MIT-licensed and scores 87/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
- Is ipcam-mcp still maintained?
- The repository was last updated 6 days ago, so ipcam-mcp is actively maintained.
Skill content
View source on GitHubipcam-mcp
<!-- mcp-name: io.github.aidev-iprokszin/ipcam-mcp -->An MCP server that lets Claude Code, or another MCP client, see through and control the IP cameras on your local network (ONVIF + RTSP): snapshots, video clips, pan/tilt and change detection, all handled locally.
Any ONVIF camera should work. See known cameras for what has been tested.
What you can ask
Once a camera is added, ask in plain words, for example:
- "Take a snapshot of the front door."
- "Save this view as door-closed." Later: "Does the front door still look like door-closed?"
- "Did anything change at the front door since the last snapshot?"
- "Pan the office camera a bit to the left."
- "Record 10 seconds from the garage."
- "Show me what the garage cameras captured yesterday."
Prerequisites
- Linux or WSL2, on the same network as the camera
- uv on your
PATH(it installs Python 3.13+ and the dependencies for you) - ffmpeg:
sudo apt install ffmpeg - An ONVIF-capable IP camera with a local camera account (username and password), with ONVIF and RTSP enabled
- Claude Code or another MCP client that supports stdio servers
Install
Pick one. All of them run the ipcam-mcp package from PyPI and need ffmpeg and, except pipx, uv (see Prerequisites).
-
Claude Code plugin (recommended for Claude Code):
/plugin install ipcam --marketplace aidev-iprokszin/ipcam-mcpOn Claude Code older than 2.1.275, run
/plugin marketplace add aidev-iprokszin/ipcam-mcp, then/plugin install ipcam@ipcam-mcp. The plugin runs the server from a lockfile (every dependency pinned with hashes). Update it withclaude plugin update ipcam@ipcam-mcp, which also updates the server; a new server version reaches the plugin shortly after its PyPI release. -
Claude Code, without the plugin:
claude mcp add ipcam -s user -- uvx ipcam-mcp -
Other MCP clients (Cursor, VS Code, Codex CLI, …): configure the stdio command
uvx ipcam-mcp. See Other MCP clients. -
A permanent
ipcam-mcpcommand:uv tool install ipcam-mcp(orpipx install ipcam-mcp), then configure your client to runipcam-mcp. Update withuv tool upgrade ipcam-mcp(orpipx upgrade ipcam-mcp).
Then add your first camera.
Add your first camera
- Camera account: create or find the camera's local ONVIF/RTSP account (username and password). This is not your cloud login. See known cameras for where your brand hides it.
- IP address: look it up in the camera's app or your router. Network discovery is not supported (WSL2 NAT blocks it).
- Ask Claude to add it, for example "add my camera at 192.0.2.10 as front-door". Claude
calls
add_camera, which asks for the username and password in a login form, so they never pass through the model. Before you type the password, check that the IP in the form is your camera's. Without a config file the server starts with no cameras;add_cameracreatescameras.tomland the camera's secrets file. See Managing cameras for the other management tools and for clients that cannot show forms.
Or configure it by hand
- Secrets file (outside any repo), e.g.
~/.config/ipcam-mcp/secrets.env:
The host is a bare IP address or hostname: noCAMERA_IP=192.168.x.y CAMERA_USER=... CAMERA_PASSWORD=...rtsp://, user name, port or path (ports go in Camera settings). Make it readable only by you:chmod 600 <file>. The server warns at startup if other users can read it. - Camera config: copy
cameras.example.tomlto~/.config/ipcam-mcp/cameras.toml(or pointIPCAM_MCP_CONFIGat another path) and adjust the camera name / variable names. Only variable names go in this file, never the secrets themselves. Usetype = "onvif".
After editing cameras.toml by hand while the server runs, ask Claude to reload the cameras
(reload_cameras).
Camera settings
All optional, set in the camera's section of cameras.toml. A wrong type is reported when the
config loads.
| Setting | Type | Default |
|---|---|---|
| onvif_port | integer 1–65535 | 80 |
| rtsp_port | integer 1–65535 | the port in the camera's stream URI, else 554 (a value you set always wins) |
| invert_pan | boolean | false |
| ptz | boolean | false |
| stream_high | path starting with / | looked up over ONVIF |
| stream_low | path starting with / | looked up over ONVIF |
ptz = false means the camera has no pan/tilt tools. Set stream_high/stream_low if the
ONVIF lookup fails or picks the wrong stream. Set invert_pan = true if ptz_move left turns
the view to the right. Some brands need other values; see
known cameras.
Other MCP clients
Any client that supports stdio servers (Cursor, VS Code, Codex CLI, …) can run the same
command: uvx ipcam-mcp, or uv --directory /absolute/path/to/ipcam-mcp run ipcam-mcp for a
clone.
Claude Desktop on Windows: the server runs inside WSL2, so install uv and ffmpeg there
(see Prerequisites). In Claude Desktop, open Settings → Developer → Edit Config, add the
ipcam entry below to mcpServers and keep the rest of the file:
{
"mcpServers": {
"ipcam": {
"command": "wsl.exe",
"args": ["-e", "/home/<you>/.local/bin/uvx", "ipcam-mcp"]
}
}
}
Replace <you> with your WSL user name (which uvx in WSL shows the path). With several WSL
distributions, add "-d", "<distro>" before "-e". Then quit Claude Desktop from the tray icon
(closing the window is not enough) and start it again. The server uses the same cameras.toml
and media folder inside WSL as Claude Code there. Use Edit Config rather than a fixed path: the
Microsoft Store version keeps the file under %LOCALAPPDATA%, not %APPDATA%\Claude. For a
clone, use "args": ["-e", "/home/<you>/.local/bin/uv", "--directory", "/home/<you>/path/to/ipcam-mcp", "run", "ipcam-mcp"].
WSL2 networking
If the camera can't be reached from WSL (nc -zv <ip> 554 fails), enable mirrored networking
in %UserProfile%\.wslconfig:
[wsl2]
networkingMode=mirrored
then run wsl --shutdown from Windows.
Tools
| Tool | What it does |
|---|---|
| list_cameras(include_removed=False) | Configured cameras with their uid, scene and capabilities (camera, ptz); include_removed=true also lists removed cameras whose files are still on disk |
| camera_snapshot(camera=None, quality="high", crop=None, region=None, from_path=None) | Returns the picture (≤1280 px wide) and saves the full-size JPEG. crop=[x, y, w, h] (fractions 0–1) or region (top-left, top, top-right, left, center, right, bottom-left, bottom, bottom-right) returns a detail at full resolution. from_path reuses an earlier snapshot (for example one from scene_timeline) without contacting a camera; camera is then optional |
| camera_save_baseline(camera, name, quality="high") | Takes a snapshot and keeps it as a named reference (e.g. door-closed) |
| camera_compare(camera, baseline=None, quality="high") | Takes a snapshot and reports what changed against the camera's previous snapshot in its current scene, or against a named baseline saved in that scene |
| camera_record(camera, seconds=10, quality="high") | Records 1–60 s to an mp4, returns its path |
| ptz_move(camera, direction, amount="small") | Pan/tilt left/right/up/down by small/medium/large; waits for the move to finish and returns the final position if the camera reports one |
| ptz_list_presets(camera) | Saved positions |
| ptz_goto_preset(camera, preset) | Move to a preset by name or token |
| ptz_save_preset(camera, name, overwrite=False) | Saves the current position as a preset; refuses an existing name unless overwrite=true |
| ptz_remove_preset(camera, preset) | Deletes a preset by name or token (cannot be undone) |
| list_scenes() | Every scene with its cameras, removed cameras that watched it, the first and last capture and counts |
| scene_snapshot(scene, quality="high") | A snapshot from every camera in the scene, each labelled with the camera's name; a camera that fails reports its error without failing the others; the call fails only if every camera does |
| scene_compare(scene, baseline=None, quality="high") | camera_compare for every camera in the scene, labelled the same way |
| scene_record(scene, seconds=10, quality="high") | Records 1–60 s with every camera in the scene at the same time, labelled the same way; returns each clip's path and size. Combining the clips (split screen etc.) is up to you |
| scene_timeline(scene, start=None, end=None, limit=50) | The scene's snapshots, recordings and baselines, newest first, from every camera that watched it (removed ones too). start/end: YYYY-MM-DD (whole day) or an ISO datetime; limit 1–500 |
camera is a camera's name: its section name in cameras.toml. quality="high" uses the
camera's main stream (its resolution is set on the camera), "low" the sub stream (smaller,
quicker). Media is saved to the
camera's own folder, <media_dir>/<uid>/<camera>_<YYYYmmdd-HHMMSS>.jpg|.mp4 (media_dir
defaults to ~/.local/share/ipcam-mcp/media).
Tool hints. Every tool carries MCP annotations so clients can tell what it does:
list_cameras, list_scenes, scene_timeline and ptz_list_presets are read-only;
camera_save_baseline, ptz_save_preset, ptz_remove_preset, remove_camera and
delete_camera may replace or delete data; everything else only adds files or moves the
camera. Snapshots and recordings are not marked read-only because they save files. These are
hints for the client; the server's own safeguards (such as delete_camera's approval form)
still apply.
Compare and crop. camera_compare lists the grid cells that changed (the same names as
region) and a box around them. It compares the frames at the size the camera sends them,
so the stream resolution you set on the camera decides how small a change it can notice; a
reference of another size is scaled to the new frame. Pass a changed cell as region to camera_snapshot, ideally
with from_path set to the snapshot you just looked at, to see it closer. Baselines live in
<media_dir>/<uid>/baselines/ as a JPEG plus a small JSON file.
Managing cameras
Instead of editing cameras.toml by hand you can let Claude add, rename, remove and delete
cameras. This is on by default. To turn it off, set allow_management = false in
cameras.toml (see cameras.example.toml), then run /mcp reconnect ipcam; the flag is only
read at startup. Claude Code still asks before each tool call unless you allowlist it; if you
allowlist the camera tools, consider leaving remove_camera, delete_camera and
ptz_remove_preset off the list.
add_camera logs into whatever host it is given. Before you type a password into the login form,
check that the IP it shows is your camera's (from the camera's app or your router), not one the
agent picked up elsewhere.
| Tool | What it does | |---|---| | `add_camera(name, host, credentials_via="form", onvif_port=None, scene=No
Truncated for display — read the full file on GitHub.
Related Skills
taste-skill
116.8kAgent skills that give coding agents taste: art direction, reference images and a browser check instead of generic AI pages
Agent-Reach
112.3kGive your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
ruflo
81.9k🌊 The original agent harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, federation, vector RAG integration, and native Claude Code / Codex / Hermes and many more Integrated
headroom
79.8kCompress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.
Languages
Trust signals
From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.
