# Clawcierge

> Human fixers for AI agents. When you hit a wall, escalate to Clawcierge. We resolve problems AI can't, provide reality checks on AI advice, and handle tasks requiring human judgment, trust, or presence.

Clawcierge is a human concierge service for AI agents. When your agent hits limits — tasks requiring human trust, judgment, or physical presence — escalate to Clawcierge. A human fixer completes the task and returns control to your agent.

## Primary Interface: Chat API (OpenAI-compatible)

```
POST https://api.clawcierge.com/v1/chat/completions
Authorization: Bearer $CLAWCIERGE_API_KEY
```

Multi-turn conversational interface. Describe what you need in natural language. Mr. Wolf (our AI concierge) triages, tickets, and routes to a human fixer.

## Docs

- [Full Documentation](/llms-full.txt): Comprehensive integration guide with examples
- [Skill File](/skill.md): OpenClaw skill for automatic integration
- [OpenAPI Spec](/.well-known/openapi.json): Machine-readable API definition

## Integration Methods

- **Chat API** (primary): `POST https://api.clawcierge.com/v1/chat/completions` — OpenAI-compatible, multi-turn conversations
- **Webhook** (fallback): `POST https://api.clawcierge.com/hooks/agent` — fire-and-forget handoff requests
- **Telegram**: [@Clawcierge_bot](https://t.me/Clawcierge_bot) — direct chat or human-in-the-loop deep links

## Service Tiers

- [Fixer Direct](/tiers/fixer-direct): <2 min response, phone/video, $75+ minimum
- [Fixer Retainer](/tiers/fixer-retainer): Reserved access, all channels, $500+/mo
- [Concierge](/tiers/concierge): ~10 min queue, chat channels, $5-15/conversation
- [Callback](/tiers/callback): Next business day, async, $3-8/request

## Optional

- [Reality Check Service](/docs/reality-check): Independent verification of AI advice (anti-sycophancy)
- [Full Documentation](/llms-full.txt): Comprehensive details for deep context