← Back to Skills Marketplace
S2 Light Perception
by
MilesXiang
· GitHub ↗
· v1.0.2
· MIT-0
149
Downloads
0
Stars
0
Active Installs
3
Versions
Install in OpenClaw
/install s2-light-perception
Description
S2-SP-OS Light Radar. Real LAN UDP discovery and local API calls (Hue/Wiz) to create a 4sqm voxel light memzero. / S2-SP-OS 光之雷达。真实的局域网UDP发现与本地API调用，逆向推算4平米光...
Usage Guidance
This skill actively scans your local network and sends UDP broadcasts/SSDP multicasts to discover Hue/Wiz devices and then makes local HTTP/UDP calls to query state. That behavior is expected for its purpose but can be noisy and may violate some network policies. Before installing or running: (1) only run it on trusted/private networks or with permission from your network administrator; (2) review the included light.py yourself — it is the only code that will run and there are no hidden downloads; (3) note that 'mqtt' is advertised but not implemented in the code, and the Hue username is a placeholder (queries to Hue bridges may fail without a real user token); (4) consider whether active device discovery is acceptable for your environment. If you need the agent to avoid any network activity, do not install or run this skill.
Capability Assessment
Purpose & Capability
Name/description describe active LAN discovery and local Hue/Wiz calls; the skill only requires python3 and performs UDP/SSDP and HTTP calls to discovered devices — this matches the stated purpose. Small mismatch: the CLI and README mention 'mqtt' as a protocol, but the code implements only 'hue' and 'wiz' handlers.
Instruction Scope
SKILL.md explicitly instructs the agent to perform active network scans (UDP broadcast to port 38899 and SSDP multicast to 239.255.255.250:1900) and to make local HTTP/UDP calls — this is expected given the purpose but is network-active behavior that will generate traffic on the LAN. The skill enforces a '--consent-granted true' gate for read operations; discovery does not require consent. No instructions attempt to read unrelated local files or exfiltrate data to external endpoints.
Install Mechanism
No install spec; instruction-only with an included Python script. No downloads or archive extraction; low install risk. Execution uses only the standard library (socket, urllib).
Credentials
The skill requests no environment variables or credentials. It does include a hardcoded placeholder Hue username path ('s2-agent-user') noted in comments as a stub — this is not a secret but may cause authentication failures when querying Hue bridges. No unrelated credentials or config paths are requested.
Persistence & Privilege
The skill is not marked 'always' and does not persist or modify other skills or system settings. It performs actions only when executed and does not request elevated or persistent privileges.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install s2-light-perception - After installation, invoke the skill by name or use
/s2-light-perception - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.2
Version 1.0.2 of s2-light-perception is a major update focused on real network-based light discovery and control.
- Switched from simulated to 100% real network-active radar using true UDP Broadcast (Port 38899) and SSDP Multicast (Port 1900).
- Introduced zero-dependency, native Python implementation for device discovery and state reading (no third-party libraries).
- Clarified instructions: device scanning and state fetching now require real network communication.
- Added robustness: estimated_lux defaults safely if real network access to a device fails.
- Updated documentation to emphasize real network interaction and provide clearer sample usage.
v1.0.1
No major code changes. Documentation improvements and internationalization.
- Improved SKILL.md with clearer English/Chinese descriptions and instructions.
- Added homepage field in metadata for easy reference.
- Enhanced bilingual documentation for all sections, making usage and features more accessible for English and Chinese users.
- Streamlined feature list for better readability and user guidance.
v1.0.0
- Initial release of S2-Light-Perception, integrating real-world light sensors and smart lighting states into unified 4㎡ grid light memory (Memzero).
- Supports active local network discovery of smart lights (Philips Hue, Wiz) and state extraction (on/off, brightness, color temperature).
- Translates light device data into S2 spatial memory grid for environmental awareness.
- Includes offline calculation of lighting effects on circadian rhythm and health, with actionable suggestions.
- Provides instructions for scanning, grid authorization, and lighting insight phases.
Metadata
Frequently Asked Questions
What is S2 Light Perception?
S2-SP-OS Light Radar. Real LAN UDP discovery and local API calls (Hue/Wiz) to create a 4sqm voxel light memzero. / S2-SP-OS 光之雷达。真实的局域网UDP发现与本地API调用，逆向推算4平米光... It is an AI Agent Skill for Claude Code / OpenClaw, with 149 downloads so far.
How do I install S2 Light Perception?
Run "/install s2-light-perception" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is S2 Light Perception free?
Yes, S2 Light Perception is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does S2 Light Perception support?
S2 Light Perception is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created S2 Light Perception?
It is built and maintained by MilesXiang (@spacesq); the current version is v1.0.2.
More Skills