This project is a Skill for reCamera V2 (RV1126B) devices. It helps agents correctly use the device's HTTP/WebSocket APIs and the related SenseCraft AI cloud APIs during development, integration, and troubleshooting.
Its focus is not just listing endpoints, but packaging real executable workflows together, such as logging in to obtain a token, using the read-before-write pattern for configuration APIs, discovering devices on a local network, accessing recording storage via relay, and handling model inference and cloud conversion flows.
- GitHub：https://github.com/mjq2020/reCamera_skill
- YouTube：https://www.youtube.com/watch?v=eh7HKjYU2SI
- ClawHub：https://clawhub.ai/mjq2020/recamera
- Authentication flow: call
/system/keyfirst, then log in and capture thetokenfromSet-Cookie - Authorized request requirements: all subsequent requests must reuse
Cookie: token=<jwt> - Configuration update pattern: for APIs that support GET/POST or GET/PUT, follow the "read before write" approach
- Device discovery workflow: identify whether a device on the local network is a reCamera through
/cgi-bin/entry.cgi/system/key - Image tuning workflow: enter scene edit mode, modify parameters, and send the save command
- Storage relay workflow: create a relay, list directories, download files, and handle timeouts
- AI capabilities: model upload, inference configuration, notification output, and SenseCraft cloud model conversion
SKILL.md: the main Skill entry file, defining trigger conditions, workflows, and key constraintsAPI_REFERENCE.md: a more complete API reference, organized by modulescripts/discover_recamera.py: a LAN scanning script used to discover reCamera devices within the subnet
- You can install the Skill already published to
clawhubusing the following method:
# Install the clawhub CLI, skip if already installed
npm i -g clawhub
# Or
pnpm add -g clawhub
# Install the Skill via CLI
clawhub install recamera
# Update the Skill
clawhub install recamera