openclaw-execute

name: openclaw_execute description: Execute Solaris analysis from OpenClaw using the deterministic pipeline endpoint.

OpenClaw Execute Skill

Purpose

Use this skill when an OpenClaw agent has explicit coordinates and needs a full, reliable Solaris run in one call.

Endpoint

• POST http://localhost:8000/openclaw/execute

Request Body

{
  "message": "Analyze energy demand and generate deployment plan",
  "lat": -1.286,
  "lon": 36.817,
  "households": 150,
  "horizon_days": 30,
  "usage_profile": "mixed",
  "thread_id": "optional-session-id"
}

Headers

• Content-Type: application/json
• x-api-key: <SOLARIS_API_TOKEN> (if auth enabled)

Success Output To Surface

• run_id
• loc_id
• outputs.demand_forecast
• outputs.scenario_set.primary
• outputs.impact_metrics
• outputs.quality
• runtime.agent_steps
• satellite (preview/NDVI/NDWI payload)

Agent Behavior Rules

1. Never invent coordinates. If missing, delegate to a geocode skill first.
2. Always return run_id and loc_id with a concise numeric summary.
3. If status is degraded, include quality_flags and assumptions in response.
4. If API fails, return exact error string and stop.