lc_scene_analysis_http

What This Skill Does

The LC Scene Analysis HTTP skill is an advanced artificial intelligence tool designed specifically for construction site safety monitoring. It leverages specialized computer vision models to process static images, automatically identifying potential hazards such as improper safety gear usage, structural risks, or site management violations. By utilizing a modular interface architecture, it provides flexible analysis, from standard hazard identification to raw data collection and high-precision cross-validation. This ensures that construction safety officers receive both quick insights and verifiable data when needed.

Installation

To install this skill, use the following command in your terminal: clawhub install openclaw/skills/skills/chenqu108/lc-scene-analysis-http

Once installed, ensure that your OpenClaw environment contains the lc_scene_http:default authentication profile. This profile must include the api_base, api_key, flow_id, and optionally an algorithm_id to function correctly.

Use Cases

• Standard Safety Inspections: Rapidly scanning images to generate a comprehensive report of safety hazards detected on a job site.
• Structured Data Export: When site management systems require raw JSON outputs for database integration or long-term record-keeping.
• Verification Protocols: Ensuring high-stakes safety decisions are supported by secondary analysis models (cross-validation) to minimize false positives.
• Compliance Reporting: Supplying documented evidence of site conditions for audit purposes.

Example Prompts

1. "Analyze this image of our excavation zone and tell me what safety hazards are present: [URL]"
2. "I need a detailed, structured JSON result for the scene captured in this image: [URL]"
3. "Please perform a secondary cross-validation on this image to confirm the presence of helmet violations: [URL]"

Tips & Limitations

• Focus on Still Images: This skill is built strictly for static image analysis. Do not attempt to use it for live video streams or real-time camera feeds.
• Configuration Overrides: While the skill relies on the default profile, you can override the flow_id and algorithm_id dynamically if your current task requires a specific diagnostic configuration.
• Data Privacy: Avoid sharing sensitive site information that might contain non-relevant personal data unless necessary. The skill handles sensitive authentication tokens internally and will never expose them in chat outputs.
• Input Requirements: Always ensure you provide a valid, publicly accessible image URL. If no image is detected, the skill will inform you that it lacks the necessary data to perform an analysis.