← Back to Skills Marketplace
160
Downloads
1
Stars
0
Active Installs
3
Versions
Install in OpenClaw
/install camera-capture
Description
Control the local computer's built-in camera for capturing photos and basic operations.
README (SKILL.md)
Camera Control Skill
Control the local computer's built-in camera for capturing photos and basic operations.
When to Use
- User wants to take a photo using the computer camera
- User asks to capture an image from the webcam
- User wants to check the camera view
- Commands like: "拍照", "打开摄像头", "拍张照片", "camera", "webcam", "take a photo"
Tools Required
- Python 3 with OpenCV (
cv2) - Standard library only (no extra deps beyond opencv-python)
Usage
Take a Photo
Capture a single frame from the default camera:
python3 skills/camera/scripts/camera.py capture --output "~/.openclaw/workspace/captures/photo_$(date +%Y%m%d_%H%M%S).jpg"
Or use the wrapper for default location:
python3 skills/camera/scripts/capture.py
This saves to ~/.openclaw/workspace/captures/ with timestamp.
List Available Cameras
python3 skills/camera/scripts/camera.py list
Test Camera (preview window, auto-close after 3 seconds)
python3 skills/camera/scripts/camera.py preview --duration 3
Output
- Photo path is printed to stdout
- Use this path with the
imagetool to analyze the captured photo - Photos are saved to
captures/directory by default
Examples
User: "拍张照片"
Action: Run capture script, then use image tool to show/analyze the photo.
User: "看看摄像头画面" Action: Run preview script for 3 seconds to verify camera works.
Notes
- Default camera index is 0 (built-in webcam)
- If you have multiple cameras, use
--camera 1,--camera 2, etc. - On Windows, the script auto-detects and works with DirectShow backend
- Captures directory is auto-created if it doesn't exist
Usage Guidance
This skill is internally coherent: it runs local Python scripts using OpenCV to access your webcam and saves images under ~/.openclaw/workspace/captures/. It does not request credentials or show any network/exfiltration code in the bundle. Before installing, consider: (1) Install opencv-python in a controlled environment and review the scripts yourself (they are short and readable). (2) Be aware that camera access is privacy-sensitive—only install if you trust the agent and its ability to invoke skills; consider disabling autonomous model invocation if you want explicit prompts before camera use. (3) You can test by running the capture script manually and inspect saved photos; remove the captures directory or uninstall the skill to revoke local artifacts.
Capability Assessment
Purpose & Capability
Name/description match the included scripts and declared requirements: python3 + OpenCV. The scripts implement list, capture, and preview and only require local camera access and a local captures directory.
Instruction Scope
SKILL.md instructs running the provided Python scripts to list, preview, or capture photos; instructions reference only the captures directory and the image tool for local analysis. There are no instructions to read unrelated files, environment variables, or to contact external endpoints.
Install Mechanism
No install spec is provided (instruction-only). Dependencies are limited to opencv-python (pip), and requirements.txt reflects that. There are no external download URLs or archive extracts in the skill.
Credentials
The skill requests no environment variables or credentials. It writes captured images to ~/.openclaw/workspace/captures/ (created if missing), which is reasonable for this purpose but is sensitive data (local images).
Persistence & Privilege
always: false and no system-wide config changes. However, model invocation is allowed (platform default), so an agent with autonomous capability could trigger the camera—this is a privacy consideration rather than a code incoherence.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install camera-capture - After installation, invoke the skill by name or use
/camera-capture - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.2
Updated version in SKILL.md and added MIT license
v1.0.1
Added MIT license with attribution requirement
v1.0.0
Initial release - camera capture, list, and preview
Metadata
Frequently Asked Questions
What is Camera Capture?
Control the local computer's built-in camera for capturing photos and basic operations. It is an AI Agent Skill for Claude Code / OpenClaw, with 160 downloads so far.
How do I install Camera Capture?
Run "/install camera-capture" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Camera Capture free?
Yes, Camera Capture is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Camera Capture support?
Camera Capture is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Camera Capture?
It is built and maintained by dddjj0 (@dddjj0); the current version is v1.0.2.
More Skills