CallChatSyn MCP Server
# CallChatSyn MCP server
Let an AI assistant (Claude, Cursor, any MCP client) answer customer questions and book appointments for a small business, using that business's own FAQs and opening hours on [CallChatSyn](https://callchatsyn.com/developers).
**Free for the first 1,000 businesses** (founder offer: 100 API calls a day, 2,000 a month).
## Tools
| Tool | What it does |
|---|---|
| `answer_customer_question` | Answers from the business's FAQs and order data (English/Chinese); flags booking and "talk to a person" requests |
| `list_open_times` | Next open appointment times in the business's time zone, plus services and locations |
| `book_appointment` | Books one of those times (only offered times; the business is notified) |
## Setup
1. Create a CallChatSyn account, add FAQs and opening hours, then Dashboard → Developers → claim founder access → create an API key.
2. Add to your MCP client config:
```json
{
"mcpServers": {
"callchatsyn": {
"command": "npx",
"args": ["-y", "github:oneappsworld/callchatsyn-mcp"],
"env": { "CALLCHATSYN_API_KEY": "ccs_live_..." }
}
}
}
```
Keep the key private; it acts for one business.
## Notes
- Answers are rule-based FAQ matching (fast, predictable), not free-form generation.
- API docs and OpenAPI spec: https://callchatsyn.com/developers · https://callchatsyn.com/openapi.json
MIT licensed.
TDQS
Scored across 3 tools
The three tools have distinct core purposes: answering informational queries, listing available times, and booking appointments. However, answer_customer_question returns an 'appointment' intent that might tempt an agent to use it for scheduling queries instead of list_open_times, creating a mild boundary overlap.
All tool names use snake_case and follow a verb-first pattern (answer_customer_question, list_open_times, book_appointment). The convention is consistent and predictable.
Three tools are well-scoped for a focused customer chat and appointment booking server; each tool serves a clear, non-redundant role and the set avoids bloat.
The surface covers answering FAQs/orders and booking appointments, but lacks tools for canceling or rescheduling appointments and for actually performing a human handoff. These are notable gaps for a customer service lifecycle.