WhatsApp CRM Support Desk
Suggested by runchr · today
Draft — a suggested stack, not yet verified by a member.
Self-hosted CRM plus support desk that answers WhatsApp buyers and logs them.
What they did
DeskcommCRM owns the contact record and WhatsApp conversation, so every inbound message already has a lead attached. Chatwoot receives the same thread as a support conversation and drafts a reply using its AI agent. MaxKB is queried first for product and policy answers, so the draft cites your own knowledge base instead of improvising, and the resolved reply is written back to the CRM record.
How it went
You get answered WhatsApp inquiries with a grounded draft, a logged CRM contact, and a closed support conversation.
Worth knowing
Two systems of record for the same conversation drift unless you pick one as the source of truth and sync on a schedule.
Wiring
Connect DeskcommCRM's MCP endpoint to the agent so it can read contacts and WhatsApp threads. Point Chatwoot's API at the same agent for conversation and reply actions. Query MaxKB over its API for grounded answers before composing any reply.
Try it yourself
A customer just asked on WhatsApp whether our annual plan includes onboarding. Pull their CRM record, check our docs for the answer, draft a reply, and log it.
Tools in this stack
Run this stack
Install the stack into this project. It writes AGENTS.md and the manifest, merges the MCP servers into .mcp.json, and — with --yes — runs the tool installs.
npx runchr install runchr/whatsapp-crm-support-deskcurl -s https://api.runchr.com/api/recipes/runchr/whatsapp-crm-support-desk/manifestSystem prompt
You are a front-line support agent for a one-person company. You answer customer messages only from the knowledge base and the CRM record, and you never invent pricing, policy, or delivery promises.
The manifest is also at api.runchr.com/api/recipes/runchr/whatsapp-crm-support-desk/manifest.