Introduction

Learn how Weavz gives your product a secure integration and MCP layer for agents, copilots, and SaaS workflows.

Introduction

Weavz is the integration and MCP infrastructure layer for modern software teams. It lets your product connect to third-party apps, manage credentials, run actions, receive events, and expose those capabilities to AI agents without building every connector, auth flow, and execution surface yourself.

Why Weavz

Modern products need more than a connector catalog. Your users expect SaaS apps, internal tools, copilots, automations, and agents to work against real business systems with the right tenant and end-user context. Weavz gives you that connective layer: auth, scoping, actions, triggers, MCP servers, and SDKs in one place.

  • 500+ pre-built integrations — Slack, GitHub, Google Sheets, Notion, Salesforce, and more
  • Setup templates — create a useful workspace, curated integrations, and an OAuth MCP server from a guided starter pack
  • MCP-native agent access — expose approved tools to Claude, Cursor, ChatGPT, and other MCP-compatible clients, with OAuth as the default for per-user workspaces
  • Code Mode MCP — let agents use a workspace-aware coding surface for richer automation and file workflows
  • Human Gates approvals — pause sensitive API, SDK, MCP, playground, or trigger execution until a reviewer decides
  • Secure credential management — OAuth2, API keys, encrypted storage, and tenant-aware connection resolution
  • Precise scoping — organizations, workspaces, end users, aliases, partials, and per-user connection strategies
  • Developer-friendly surfaces — Dashboard, REST API, TypeScript SDK, and Python SDK

Core Concepts

Integrations

An integration represents a third-party service (e.g., Slack, GitHub, Google Sheets). Each integration defines its authentication method, available actions, and triggers. Weavz ships with 500+ integrations plus built-in storage and key-value store integrations.

Workspace Integrations

A workspace integration is a configured instance of an integration inside a workspace. It gives that instance an alias, connection strategy, enabled actions, and optional settings such as Storage/KV persistence or Advanced Code sandbox policy. Agents and SDK callers should usually target workspace integrations rather than raw connections.

Learn more about workspace integrations

Connections

A connection is an authenticated credential for a specific integration. When a user connects their Slack account via OAuth2, or provides an OpenAI API key, that creates a connection. Connections are encrypted at rest and scoped to your organization, workspace, or individual user.

Learn more about connections

End Users

End users map your app's users to Weavz. They let per-user workspace integrations resolve the right connection and keep built-in state scoped to the person using the agent or workflow.

Learn more about end users

Actions

Actions are operations you can execute against a third-party service — sending a Slack message, creating a GitHub issue, or reading a Google Sheet. Each action has a defined input schema and uses a connection for authentication.

Learn more about actions

Triggers

Triggers let you receive events from third-party services. When a new Slack message arrives, a GitHub PR is opened, or a Google Sheet row is added, Weavz delivers the event to your callback URL.

Learn more about triggers

MCP Servers

MCP (Model Context Protocol) servers expose your integrations as tools that AI agents can discover and use. Create an MCP server, add workspace tools, choose Tool Mode or Code Mode, and connect it to Claude, Cursor, ChatGPT, or any MCP-compatible client.

Learn more about MCP servers

Partials and Human Gates

Input partials save defaults or locked parameters for actions and MCP tools. Human Gates pause sensitive API, SDK, MCP, Playground, or trigger execution until a reviewer approves or rejects it.

Learn more about input partials and Human Gates

First-Day Path

If you are evaluating from the dashboard, use this order:

  1. Open Templates and create a starter workspace when one matches your use case
  2. Add or review workspace integrations, aliases, connection strategies, and built-in tools such as Storage, KV Store, or Advanced Code
  3. Connect the accounts and end users required by that workspace
  4. Add input partials and Human Gates for external sends, destructive writes, spend, exports, or end-user credentials
  5. Review the OAuth MCP server and keep Code Mode enabled for broad agent workflows
  6. Validate the flow in Playground
  7. Create scoped API keys and hand the tested API, SDK, or MCP setup to your app or agent

How It Works

graph LR
    A[Your App] --> B[Weavz API]
    B --> C[Third-Party Service]
    D[MCP Clients] --> B
  1. Create a workspace — scope the product, customer, project, or agent environment
  2. Configure workspace integrations — set aliases, connection strategy, enabled actions, and built-in persistence or sandbox settings
  3. Create connections and end users — authenticate accounts and resolve per-user context
  4. Add guardrails — use partials and Human Gates before exposing sensitive actions
  5. Execute actions and listen for events — call third-party APIs or enable triggers inside the workspace
  6. Serve AI agents — create MCP servers with OAuth access so AI tools can safely use workspace capabilities

Need on-premise deployment, dedicated service accounts, or custom SLAs? Contact our sales team for enterprise options.

Next Steps

  • Setup Templates — create a prepared workspace and OAuth MCP server from the dashboard
  • Installation — set up your API key and install the SDK
  • Quick Start — build your first integration end-to-end
  • Human Gates — add approval review to sensitive execution paths
  • Playground — validate actions, MCP, triggers, partials, and approvals before launch
  • Integrations — browse available integrations