Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Detects the primary emotion in a text string and returns structured metadata: emotion type, intensity, valence, confidence score, and a recommended response strategy. Includes a safe_t_flag that triggers when critical-intensity distress signals like grief or despair are detected. Accepts English and Japanese input.
Returns a ready-to-act response_strategy alongside raw emotion data, so agents skip the step of inferring how to respond and act on a clear directive instead.
When to use it
- Routing user messages to empathetic vs. task-focused agent responses
- Flagging high-distress support tickets for human escalation
- Adjusting chatbot tone based on detected user frustration
- Filtering incoming feedback by emotional valence before summarizing
- Triggering crisis resources when a user message shows critical grief or fear
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: 'I've been feeling completely hopeless lately, nothing seems to matter anymore'
- 1Send POST request to the x402 endpoint with the user's text and language set to 'en'
- 2Receive response with primary_emotion: grief, intensity: critical, confidence: 0.95
- 3Check safe_t_flag — value is true
- 4Halt normal conversation flow
- 5Present user with crisis support resources (Crisis Text Line, いのちの電話)
Agent surfaces emergency mental health resources instead of continuing the task flow
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.