Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Intercepts messages from configured contacts, strips emotional content, and extracts only the factual requests and deadlines. Supports email and iMessage channels with configurable summarization modes. Drafts neutral responses for user approval before sending.
Instead of manually parsing emotional messages to find the one actionable item buried inside, the mediator surfaces only what requires a response.
When to use it
- Filtering emotional emails from an ex-partner during co-parenting
- Handling messages from a difficult client without reading the full tirade
- Getting a calm summary of a heated family group chat
- Drafting a measured reply to a passive-aggressive coworker
- Reviewing archived originals when needed for legal context
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Process new email from ex@email.com
- 1Check sender against configured contacts in mediator.yaml
- 2Fetch full email content via Gmail Pub/Sub
- 3Run email through LLM to extract actionable items and strip emotional language
- 4Archive original email with 'Mediator/Raw' label and mark as read
- 5Send facts-only summary and draft response to configured notify channel
Summary: 'Request: Pick up kids at 3pm Saturday' with suggested reply 'Confirmed. I'll pick up the kids at 3pm on Saturday.'
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.