flowcastle-mcp
Build, edit, and deploy Telegram bots on FlowCastle's hosted visual flow platform: 16 tools to read workspace context, learn the flow format, validate and apply flow edits, and manage broadcasts. Ships with 23 ready-to-use bot templates and runs as a stdio bridge (mcp-remote) to the hosted Streamable HTTP endpoint.
README
Telegram bot templates + MCP server for AI bot development
Ready-to-use Telegram bot templates and an MCP server — build, edit, and deploy Telegram bots from Claude Code, Codex, Cursor, or any MCP-capable AI agent. No SDK boilerplate: describe the bot, and your agent assembles the flow.

FlowCastle is a visual workflow builder for Telegram bots and web chat: flows, triggers, conditions, AI steps, payments (Stripe / YooKassa / Telegram Stars), broadcasts, drip sequences, and integrations (Gmail, Google Sheets, Notion, OpenAI, …) — with a REST API and an MCP server, so bot development works straight from your AI coding agent. Every block an agent creates through MCP is the same block you can drag, wire, and edit visually.
This repo is for developers. It contains:
| Path | What it is |
|---|---|
templates/ |
23 production bot templates as JSON — the same source that powers the template gallery and the MCP get_flow_example tool |
mcp/ |
MCP server reference: endpoint, auth, all 16 tools, and a snapshot of the flow-action schema used by apply_actions |
Build a bot from your AI agent (MCP)
FlowCastle exposes an MCP server over Streamable HTTP:
POST https://api.flowcastle.ai/api/mcp
Authorization: Bearer <your-application-api-key> # starts with app_
Get a key: sign up → create an application → API page → MCP tab.
Claude Code
export FLOWCASTLE_MCP_TOKEN='<your-application-api-key>'
claude mcp add --transport http flowcastle https://api.flowcastle.ai/api/mcp \
--header "Authorization: Bearer $FLOWCASTLE_MCP_TOKEN"
Codex
codex mcp add flowcastle \
--url https://api.flowcastle.ai/api/mcp \
--bearer-token-env-var FLOWCASTLE_MCP_TOKEN
Cursor / any MCP client (JSON config)
{
"mcpServers": {
"flowcastle": {
"type": "http",
"url": "https://api.flowcastle.ai/api/mcp",
"headers": {
"Authorization": "Bearer <your-application-api-key>"
}
}
}
}
Then just ask your agent things like:
"Look at my FlowCastle workspace and add a lead-qualification flow: greet on /start, ask three button questions, save answers to contact variables, and notify me."
The agent reads your workspace with get_workspace_summary / get_flow_context, drafts a batch of flow-builder actions, checks it with validate_actions, and applies it with apply_actions. See mcp/README.md for the full tool reference.
Templates
Each JSON file in templates/ is a complete bot: metadata (title, summary, use-cases, suggested variables, required modules) plus a schemaExample.actions array — an ordered batch of create_flow / create_block / create_link / create_variable operations in the exact format apply_actions accepts.
Three ways to use them:
- One-click install from the gallery: flowcastle.ai/templates
- Via MCP — ask your agent to
search_flow_examples/get_flow_exampleand apply one (optionally customized) to your workspace - As reference — study how real flows wire triggers, conditions, wait-for-reply steps, payments, and AI blocks
Full index and format description: templates/README.md.
REST Public API
Application-scoped REST API for contacts, contact variables, goal events, flow sends, and broadcasts:
- Base URL:
https://api.flowcastle.ai/api/public/v1 - Interactive docs (Swagger): https://api.flowcastle.ai/api/public/docs
- OpenAPI spec: https://api.flowcastle.ai/api/public/openapi.json
Auth uses the same app_… application key (Authorization: Bearer or X-API-Key).
Guides
Step-by-step builds on the FlowCastle blog:
- Build a Lead Bot in Telegram in 10 Minutes
- Build an AI Support Bot for Telegram
- I Replaced My Morning Apps With One Telegram Bot
Links
- Product: https://flowcastle.ai
- Dashboard: https://dashboard.flowcastle.ai
- Template gallery: https://flowcastle.ai/templates
- API docs: https://api.flowcastle.ai/api/public/docs
Contributing
Found a bug in a template, or built a flow worth sharing? Open an issue or PR — templates are plain JSON, and validate_actions (via MCP) will tell you if a batch is well-formed.
License
The contents of this repository are licensed under the MIT License.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.