Weavz

Weavz is a governed app access layer for AI agents and SaaS products, connecting apps with scoped credentials, approvals, files, state, and MCP access.

Desarrollo Agentes IA Herramientas IA Dev No-Code y Low-Code

Visitar Sitio Web

What Weavz does

Weavz is a governed app access layer for AI agents and SaaS products. It sits between AI clients or backend workflows and third-party business apps, so teams can connect accounts, execute actions, receive events, and preserve scoped state without building every connector or approval flow themselves.

The product is centered on workspaces. Each workspace can mount integrations such as Slack, Gmail, HubSpot, Salesforce, or internal APIs, then expose approved tools to agents through MCP, SDKs, or direct action calls. The source describes Weavz as useful for chat-based assistants, copilots, automations, and embedded workflow products that need credentials, approvals, files, state, and audit in one controlled environment.

Features

Scoped workspaces and tenant boundaries

Create workspaces for customers, teams, projects, or agents, then attach integrations with aliases, connection strategy, and scoped permissions so each environment stays separated.

MCP access with Tool Mode and Code Mode

Expose app actions through hosted MCP servers and Code Mode or Tool Mode, allowing agents to discover typed APIs, run parallel calls, and execute multi-step workflows.

Approvals and action guardrails

Use Human Gates and input partials to review sensitive actions, lock defaults, prefill fields, or hide preset values from tool schemas before execution.

Filesystem and State KV primitives

Persist files and key-value state at workspace, user, or external-ID scope so agents can keep artifacts and checkpoints across runs.

Multiple control surfaces

Provision customers, workspaces, connect links, MCP servers, and scoped tokens through API, TypeScript SDK, Python SDK, CLI, or dashboard.

Credential and identity management

Connect approved accounts through hosted links and manage org- or workspace-scoped API keys for customer-facing automation.

Common use cases

Embed agent workflows in a SaaS product
Give a customer-facing AI feature access to the customer’s own connected apps while keeping credentials, state, and permissions isolated per workspace.
Use an AI client with approved app access
Connect ChatGPT, Claude, Cursor, or another MCP client to a governed workspace so the agent can act on approved business tools instead of raw credentials.
Run repeatable multi-step agent jobs
Create a workspace with files, state, Human Gates, and aliases for recurring tasks such as support follow-up, research, or internal operations.
Automate onboarding and tenant setup
Provision workspaces, end users, connect links, and scoped tokens from your backend when building an AI-native application.
Publish governed tools for agents and frameworks
Expose selected actions through MCP or direct APIs when you need a controlled integration surface for internal tools, coders, or agent frameworks.

Pros and Cons

Pros

Combines governed credentials, actions, triggers, files, state, and approvals in one workspace model.
Supports both product builders and AI-client users through API/SDKs, MCP servers, CLI, dashboard, and remote MCP clients.
Offers scoped primitives such as end users, workspace aliases, Human Gates, partials, and tenant-aware connection resolution.
Provides a free plan for evaluation and clearly published paid tiers for production use.

Cons

The source does not provide full implementation details for every supported integration or MCP setup path.
Some capabilities, including browser control and audit flow, are described at a high level rather than with step-by-step examples.
The pricing page shows capacity-based limits and prepaid add-on packs, so larger workloads may require upgrading or purchasing extra capacity.

FAQ

How do you get started with Weavz?

Weavz is designed to be used through the dashboard, API/SDKs, MCP servers, CLI, or directly from supported AI clients. The docs describe a first-day flow that starts with creating a workspace, adding integrations and end users, setting partials or Human Gates, and then testing the flow in Playground.

What apps can Weavz connect to?

The source lists Slack, GitHub, Google Sheets, Notion, Salesforce, Gmail, Google Calendar, HubSpot, Airtable, Asana, Jira, Dropbox, Shopify, Stripe, Zendesk, and many others. The homepage also says Weavz connects to 1,000+ apps, but the exact supported list is broader than the examples shown here.

Can Weavz be used from AI clients like ChatGPT or Claude?

Yes. The homepage and docs say Weavz supports ChatGPT, Claude, Cursor, Codex, Claude Code, VS Code, Cline, Gemini CLI, OpenCode, and other MCP-compatible clients, along with custom remote MCP clients.

Is there a free plan?

The pricing page shows a Free plan and paid plans for Pro, Team, and Scale, plus an Enterprise option with custom pricing and contract-based capacity. The Free plan is intended for exploring agents, actions, and state, while paid tiers increase capacity and included primitives.

How does Weavz handle approvals and access control?

Weavz is built around governed access: connections are scoped to organizations, workspaces, or end users, and Human Gates can pause sensitive API, SDK, MCP, Playground, or trigger actions until a reviewer approves them.

Quick Facts

Category: Governed app access for AI agents
Primary users: AI teams, SaaS builders, and product teams
Access methods: API, TypeScript SDK, Python SDK, CLI, dashboard, and MCP
Platform focus: Workspace-based app access with approvals, files, state, and audit
Website: weavz.io
Pricing: Free plan plus Pro, Team, Scale, and Enterprise