AI Scheduling Platform

Mission

Build a reusable OpenClaw-managed scheduling platform that standardizes self-hosted booking infrastructure for client brands.

This repository is the scheduling counterpart to the shared chatbot platform. It exists so booking can be provisioned, configured, updated, and audited once through OpenClaw, while each client gets an isolated deployment bundle.

Architecture Direction

The operating model in this repository is:

• separate from the chatbot runtime repo
• built around self-hosted Cal.com as an upstream service
• templated for per-client deployments from one control-plane repo
• OpenClaw-managed for provisioning, rollout, and lifecycle
• Postgres-backed with one isolated Cal.com deployment per client
• client-configurable for brand, email owner, booking event, and public URL

Critical design choice:

• this repo does not try to make the open-source edition of Cal.com multi-tenant
• instead, OpenClaw uses one reusable template and provisions one isolated instance per client
• that keeps us compatible with the open-source path and avoids feature/licensing drift

Phased Rollout

1. Lock the repository architecture and templating model.
2. Define the tenant config contract and deployment bundle format.
3. Create the first reusable client bundle for Wellness by Madeline.
4. Deploy the first self-hosted Cal.com instance on its own subdomain.
5. Wire the booking URL back into the chatbot platform and OpenClaw lane memory.
6. Reuse the same provisioning path for future clients.

Repository Shape

• config/templates: reusable client template contracts
• config/clients: concrete client scheduling configs
• deployment/templates: rendered Docker and env templates for upstream Cal.com
• deployments/generated: generated per-client deployment bundles
• openclaw: operator memory for the OpenClaw control plane
• docs and architecture: decision records and operating guidance
• src/scripts: repeatable client bootstrap and bundle-render scripts

Quick Start

npm install
npm run check
npm run client:init -- wellness-by-madeline
npm run client:render -- wellness-by-madeline

The render command will generate a deployment bundle at:

deployments/generated/<client-slug>/

That bundle includes:

• .env.template
• docker-compose.override.yaml
• INSTANCE_README.md

Current Direction

The first live client target is Wellness by Madeline:

• public scheduler label: Wellness by Madeline
• public scheduler subdomain target: book.wellnessbymadeline.com
• owner email: wellnessbymadeline@gmail.com
• booking purpose: 1:1 consultation handoff from The Wellness Companion

Operator Docs

• Architecture Index
• Current Phase
• Platform Direction
• Tenant Isolation Decision
• Cal.com Service Blueprint
• Control Plane Memory