Capture browser API traffic
A page-context script patches fetch and XMLHttpRequest so the extension can observe JSON API calls made by the site, including calls to third-party domains, while ignoring static assets and trackers.
Web2MCP
Web2MCP is a browser extension that records a website’s JSON API calls and turns them into native WebMCP tools for in-browser AI agents. It helps users and teams work with existing web app APIs without manual endpoint wiring or schema files.
What Web2MCP is
Web2MCP is a browser extension that records the JSON APIs a website calls and turns them into native WebMCP tools for in-browser AI agents. It is designed to let an agent work with a site’s real API surface without per-endpoint setup, schema files, or manual wiring.
The product watches requests in page context, deduplicates them into endpoint templates, infers schemas from observed traffic, and injects the resulting tools through the browser’s modelContext API. The result is a local workflow where browsing a site builds the tool set that an agent can later call.
Core capabilities
Template endpoints and infer schemas
Repeated requests are deduplicated into endpoint templates such as /users/{id}, and the extension infers JSON Schemas and required fields from what it has observed.
Expose tools through WebMCP
Each endpoint is registered as a native WebMCP tool through document.modelContext, so in-browser agents can discover and call it without manual wiring.
Replay requests in context
Calls are replayed in page context, which lets recorded cookies and captured auth headers carry through when the tool is invoked.
Keep one tool pool per website
Tools are pooled by website domain, so subpages and subdomains share one growing tool set as the user moves around a site.
Provide visibility and control
The interface includes a toolbar badge, a popup list of tools, and a per-site disable option for injection control.
Practical ways to use Web2MCP
Cookie-authenticated web apps
For internal admin panels, dashboards, and other cookie-authenticated apps, Web2MCP can capture the requests the app already makes and expose them as tools for an in-browser agent.
First-party product workflows
Teams working on products they control can browse their own app, let the extension learn the endpoints organically, and then let an agent call those tools without separate integration work.
Read endpoints and token-based requests
For read-heavy sites and endpoints that send static bearer tokens or session values in request parameters, Web2MCP can replay the observed requests with the captured context.
Multi-page site exploration
As users move across pages and subdomains of the same site, the shared domain pool lets the tool set accumulate rather than resetting on each page.
Pros and Cons
Pros
- Captures JSON API traffic as you browse, including calls to the site’s own backend and third-party APIs.
- Converts observed requests into endpoint templates and inferred schemas instead of requiring manual endpoint definitions.
- Registers tools natively through the browser’s WebMCP/modelContext flow, so in-browser agents can discover them directly.
- Keeps tools pooled by website domain, which helps tool coverage accumulate as users navigate an app.
- Runs locally with no account or server according to the site, and stores data on the user’s machine.
Cons
- It cannot reliably replay some hardened patterns, including service-worker-injected auth that the page never exposes, one-time anti-replay tokens, and request-body signatures recomputed for each payload.
- The source material is limited to the homepage and pricing page, so detailed plan structure and broader product scope are not confirmed here.
FAQ
What does Web2MCP do?
Web2MCP records the JSON APIs a website calls while you browse, deduplicates them into endpoint templates, and exposes them as native WebMCP tools in the browser. The site describes the flow as record, aggregate, and inject.
How is it installed and used?
The product adds a browser extension, then works as you browse normally. A toolbar badge shows how many tools have been captured, and the popup lists them. If WebMCP is not available yet, Web2MCP can still record requests until a WebMCP-capable browser is present.
Where does Web2MCP fit best, and where does it not work well?
Web2MCP is intended for cookie-authenticated apps, internal tools, dashboards, many B2B SaaS apps, static bearer-token requests, read endpoints, and sites you control. It is described as limited for service-worker-injected auth, one-time anti-replay tokens, and request-body signatures recomputed per payload.
Does Web2MCP send data to a hosted service?
The site says Web2MCP is open source, runs entirely on the user’s machine, stores everything locally, and requires no account or server.
Quick Facts
- Category
- Developer Tool
- Product type
- Chrome extension / browser tool
- Primary use
- Turn website API traffic into agent-callable tools
- Platform
- Chrome and Edge; works with WebMCP-capable Chromium browsers
- Availability
- Add to Chrome via the Chrome Web Store; no account required
- Source domain
- web2mcp.dev
Alternativas ao Web2MCP
CreateOS Sandbox
CreateOS Sandbox is an isolated compute environment for running code and agent workloads inside Firecracker micro-VMs. It is designed for workflows that need machine-level isolation, private networking between sandboxes, and programmatic control through SDK, CLI, or MCP.
AakarDev AI
AakarDev AI helps teams manage AI provider access, project-level setups, logs, and analytics from one dashboard. It supports BYOK workflows and lists providers including OpenAI, Google Gemini, Anthropic, Groq, Mistral AI, and Perplexity AI.
Trigger.dev chat agent
Trigger.dev chat agent is a durable AI chat backend for developers building stateful conversations that can survive refreshes, crashes, and long-running turns. It connects with the AI SDK `useChat` flow and runs on managed infrastructure with no timeout on a turn.
Arduino VENTUNO Q
Arduino VENTUNO Q is an edge AI computer for AI and robotics applications. It combines AI inference and deterministic control on a single board and is designed to work with Arduino App Lab.
ByteAsk
ByteAsk is a terminal-first AI coding agent for C and C++ that edits repositories and verifies changes with the real compiler, debugger, sanitizers, and tests before showing a diff. It offers a free tier plus paid plans, with editor connectors and zero-retention handling described in the source.
Codex Plugins
Codex Plugins bundle reusable skills, app integrations, and MCP servers into workflows you can install in the Codex app or use from Codex CLI. They help extend Codex with connected-service tasks, reusable instructions, and shared team workflows.