Skip to main content
Waaru
Native MCP server

Connect AI assistants to WhatsApp through an MCP server.

Model Context Protocol is an open client-server protocol for connecting AI applications to tools, resources, and prompts. Waaru exposes governed WhatsApp operations through MCP so an approved Claude, GPT, Gemini, or custom agent can work with messages, contacts, flows, and analytics without a bespoke connector for each action.

In one line

A WhatsApp MCP server lets an MCP-compatible AI application discover and call approved WhatsApp Business tools. Waaru provides workspace-scoped tools for messages, contacts, flows, broadcasts, analytics, handoff, and knowledge operations.

Create free workspace

Platform free. AI features are paid separately.

40+Tools exposed over MCP
Per-workspaceAPI key, scoped, revocable
JSON SchemaTyped inputs for each tool
MCP-nativeDesigned for agents, not just shipped to tick a box

The problem

An MCP endpoint is only as useful as its tools and controls.

MCP standardises capability discovery and communication between clients and servers. It does not guarantee complete tool coverage, safe permissions, approval gates, or useful logs. Those implementation details decide whether an agent can operate a customer workflow responsibly.

  • Coverage: which messages, contacts, flows, broadcasts, and analytics operations are exposed?
  • Scope: what can each workspace key read or change?
  • Approval: which high-impact actions require a person?
  • Observability: can your team inspect every call, result, and failure?

What MCP-native means in Waaru.

MCP-native is product language, not an official capability defined by the protocol. In Waaru it means approved agent operations are designed as documented, workspace-scoped tools with the same underlying product rules used by the dashboard. Messaging, flow, broadcast, contact, handoff, knowledge, and analytics operations are exposed intentionally for agent use.

40+ tools covering the full WhatsApp Business surface.

  • Messaging — send_text, send_template, send_media, send_buttons, send_list
  • Contacts — upsert, tag, segment, search, fetch_record
  • Conversations — list, fetch, fetch_history, summarize, classify_intent
  • Flows — list, create_draft, publish, run, version_compare
  • Broadcasts — create, schedule, send_now, report
  • Templates — list, submit_for_approval, status
  • Knowledge — index_url, upload_pdf, search, list_indexed
  • Analytics — fetch_campaign_report, fetch_inbox_report, fetch_flow_funnel
  • Handoff — escalate, assign, resolve
  • Resolutions — list, index, edit, delete

Auth — per-workspace API keys, scoped, revocable.

Every MCP connection is authenticated with a workspace-scoped API key. Keys are revocable from the dashboard. You can issue multiple keys per workspace with different tool allowlists, such as read-only analytics for a reporting agent or a broader service scope for a customer-service agent. Access remains attributable to the key and workspace used for the connection.

Schemas and docs auto-generated, agent-readable.

Every tool ships with a JSON-schema for its arguments and a structured docstring the agent reads at discovery. Tool descriptions are reviewed by humans, tested against real agent prompts, and updated when behaviour changes. Your agent should never be surprised by what a tool returns.

Example jobs an approved agent can complete.

Waaru over MCP lets an approved agent use WhatsApp operations as tools while deterministic flows, permissions, and human approvals remain available. Example workflows include:

  • A sales agent that qualifies a lead in chat, books a selected Google Calendar or Cal.com slot, and creates an approved Zoho CRM follow-up.
  • An order-support agent that reads bounded Shopify order and fulfilment data before drafting a customer reply.
  • An internal ops agent that any team member can prompt with 'send a check-in to every customer who hasn't responded in 14 days' — and the agent runs the segmentation, drafts the message, and ships the broadcast.
  • A Gemini support agent that watches every escalated conversation, drafts the human reply, and sends only after the human approves.
FAQ

Frequently asked questions.

What is a WhatsApp MCP server?

A WhatsApp MCP server is a Model Context Protocol server that lets an AI agent operate a WhatsApp Business workspace — sending messages, managing contacts, triggering flows, reading analytics — through standardized MCP tool calls. Your Claude, GPT, or Gemini agent uses the WhatsApp tools the same way it uses any other MCP tool.

Do other WhatsApp platforms have an MCP server?

MCP is an open protocol and is available across a growing range of software products. Do not compare only whether an endpoint exists. Compare the tools exposed, workspace scopes, approval gates, audit logs, transport, documentation, and support for your intended WhatsApp journey.

What can my agent do over MCP?

Send messages, schedule broadcasts, create and publish flow drafts, look up conversations, classify intent, escalate to a human, search the knowledge base, fetch analytics, manage contacts, and edit indexed resolutions. Waaru exposes more than 40 tools across the WhatsApp workspace surface.

How does authentication work?

Per-workspace API keys, scoped by tool allowlist, revocable from the dashboard. You can issue multiple keys per workspace — a read-only key for a reporting agent, a full key for a service agent. No shared credentials.

Can I run a custom agent on my own infrastructure?

Yes. Waaru's MCP server is reachable over the standard MCP transport. You bring the agent runtime and the model. You can run the agent on your own infrastructure and use Waaru as the WhatsApp tool surface.

Does Waaru work without an external agent?

Yes — Waaru ships its own first-class AI agent that uses the same MCP tools internally. MCP is for teams that already have an AI agent and want WhatsApp as a channel. The built-in agent is for teams that want a turnkey solution.

Is the MCP server included in every plan?

Public signup is open. The platform is free. AI features are paid separately. Meta messaging charges remain separate.

Where are the schemas documented?

Every tool includes a JSON Schema and description available through MCP capability discovery. During onboarding, Waaru support also provides the current endpoint, authentication, scope, and tool reference for the workspace.

Research notes

Evidence checked for this page.

Protocol definitions on this page follow the official MCP architecture documentation and Anthropic's original announcement.

  • MCP uses a client-server architecture with capability negotiation and primitives including tools, resources, and prompts.

    Model Context Protocol · Architecture overview · Accessed 8 August 2026

  • Anthropic introduced MCP as an open standard for connecting AI assistants to data sources and business systems on 25 November 2024.

    Anthropic · Introducing the Model Context Protocol · Accessed 8 August 2026

Keep reading.

Compare

Evaluating platforms? Every claim cited, every price dated.

Ready to ship a WhatsApp flow that doesn't break?

Public signup is open. The platform is free. AI features are paid separately.

Create free workspaceSee pricing