Listed from the MCP registry; nobody on Odel publishes it. Is this yours? Claim @chatbotx
Omnichannel customer messaging, CRM contacts, flows, sequences, broadcasts, and AI automation.
One address for any MCP client. You sign in once, after adding it.

At most: only reads or looks up information.
Odel charges nothing per call. The publisher may charge for its own service.
Odel could not list its tools without signing in.
Agentic chat marketing platform built for OpenClaw, Hermes, and Claude
An alternative to ManyChat, Chatfuel, Wati, Respond, etc...
Website | Roadmap | Docs | Discord
Register
·
CLI
·
MCP
·
Public API
N8N custom node · Make.com integration
![]() | ![]() |
|---|---|
![]() | ![]() |
![]() | ![]() |
| Feature | ManyChat | ChatbotX |
|---|---|---|
| Self-Hosted | ❌ | ✅ |
| White Label | ❌ | ✅ |
| CLI & MCP | ❌ | ✅ |
| Cost Growth | ✅ | ❌ |
| Public API | Limited | ✅ |
| Omnichannel | Limited | ✅ |
| AI Agents | Limited | ✅ |
| AI Credits | Add-on | BYOK |
apps/builderTo have the project up and running, please follow the Quick Start Guide.
Self-hosting ChatbotX without developing on it? Use chatbotx-docker-compose — a standalone repo that runs the pre-built ghcr.io images with docker compose up, no local checkout of this monorepo required. The compose files in this repository (docker-compose.yml, docker-compose.dev.yml) provision local infrastructure only and expect the apps to run from source via pnpm dev.
.
|-- apps/
| |-- builder/ # Next.js web app and product builder
| |-- worker/ # background workers for chat, AI, triggers, webhooks, analytics, sequences
| |-- realtime/ # realtime server
| |-- cli/ # ChatbotX command line client
| `-- mcp-server/ # MCP server backed by public APIs
|-- integrations/
| |-- whatsapp/
| |-- messenger/
| |-- instagram/
| |-- threads/
| |-- telegram/
| |-- zalo/
| |-- tiktok/
| |-- webchat/
| |-- smtp/
| |-- openai/
| |-- google-sheets/
| `-- ... # email/CRM: mailchimp, klaviyo, active-campaign, drip, get-response, mailer-lite, moosend, sendgrid
|-- packages/
| |-- database/
| |-- ai/
| |-- auth/
| |-- business/
| |-- automated-response/
| |-- analytics/
| |-- event-bus/
| |-- mail/
| |-- imports/
| |-- flow-config/
| |-- variables/
| |-- sdk/
| |-- scheduler/
| |-- sequence-scheduler/
| |-- ui/
| `-- worker-config/
|-- docker-compose.yml
|-- pnpm-workspace.yaml
`-- turbo.json
pnpm dev # run turbo dev
pnpm build # build all packages/apps through Turborepo
pnpm lint # run Ultracite lint
pnpm fix # run Ultracite fix
pnpm check:circular # check circular dependencies
pnpm check:unused # check unused files and dependencies
Useful package-level commands:
pnpm --filter builder dev
pnpm --filter worker dev
pnpm --filter realtime dev
pnpm --filter chatbotx dev:cli
pnpm --filter chatbotx-mcp dev:mcp
pnpm --filter @chatbotx.io/database db:studio
The worker image defaults to worker standalone: core queues, schedule, and
sequence workers share one Node process so common application code is loaded
once. Use worker core for local development without schedules, or worker <name> to isolate a workload such as heavy.
worker all remains available for one process per worker. It caps each child
heap at WORKER_MAX_OLD_SPACE_MB only when it is set (there is no default);
size it so the host has enough memory for every child. The standalone process retains
the normal V8 heap limit.
DATABASE_POOL_MAX defaults to 10 connections per process. In the worker
Docker image, consolidated worker standalone and worker core modes default
to 80 (SHARD_POOL_MAX to 40); set the variables explicitly to override.
Keep replicas × (DATABASE_POOL_MAX + shards × SHARD_POOL_MAX) at or below
PostgreSQL max_connections. The consolidated process runs about 350
concurrent job slots (trigger 100, sequence 100, webhook 50, …) over these pools.
The default Docker Compose stack includes:
5432637955409000 and console on 90011025 and UI on 80258080If ChatbotX helps you avoid expensive chatbot automation tools, please give us a ⭐ on GitHub. We're also working hard to bring open-source mobile apps to iOS and Android this year
ChatbotX' Community Edition is released as open source under the MIT License and enterprise features are released under Commercial License