src ▸ site opened 0

Turn Target into MCP tools

Give your AI agent Target tools. Writ turns your Target workflows into authenticated MCP tools your agent can call, taking real actions on your own account.

Writ runs on your own accounts, with your own credentials and data, on sites you are authorized to use.
claude_desktop_config.jsonMCP · writ-target
json
{
  "mcpServers": {
    "writ-target": {
      "url": "https://api.usewrit.app/mcp/your-workspace",
      "headers": { "Authorization": "Bearer wt_8f2c...a91" }
    }
  }
}
MCP server any client authenticated
new ▸ tool minted 01
02
create

No Target MCP server yet? Build one in minutes

Record the Target task once, or describe it to an AI session, and Writ turns it into an MCP tool your agent can call, taking real actions on your own Target account.

  1. 1
    Record a Target action once

    Open the recorder and search a keyword on Target in a real browser. Writ captures the steps, login and 2FA included, into a workflow.

  2. 2
    Publish it as an MCP tool

    One click exposes the workflow as an MCP tool with a typed input schema, alongside a REST endpoint. Your agent picks it by its description.

  3. 3
    Call it from any MCP client

    Point Claude Desktop, Claude Code, Cursor, or any MCP client at your Writ server. The Target tool acts on your own account and returns structured output.

Browse supported MCP clients, see what's possible on workflows, or search for another site. Free to start. See pricing.

gap ▸ no official mcp 02
03
why

Target has no MCP server

Read-only website-to-MCP tools can summarize a Target page, but they can't do anything on it. Writ turns the Target workflows you build into MCP tools your agent calls to take real actions (search, extract, submit), authenticated, on your own Target account.

Writ runs on your own accounts, with your own credentials and data, on sites you are authorized to use.

tools ▸ actions named 03
04
named tools

Target tools your agent can call

Each tool is a real Target workflow with a typed input schema and structured output. Your agent picks the right one by its description.

{ }target.get_priceReturn price and stock for a product URL.
{ }target.check_storeCheck availability of a product at a chosen store.

Example tool schema

target.get_price.jsonschema
json
{
  "name": "target.get_price",
  "description": "Return price and stock for a product URL.",
  "input_schema": {
    "type": "object",
    "properties": {
      "query": { "type": "string", "description": "What to look up on Target." }
    },
    "required": ["query"]
  }
}
wire ▸ any client 04
05
connect

Add the server to any MCP client

Copy the config into your client and authenticate with a wt_ key. Tools appear automatically.

Your key (wt_)
json
{
  "mcpServers": {
    "writ-target": {
      "url": "https://api.usewrit.app/mcp/your-workspace",
      "headers": { "Authorization": "Bearer wt_8f2c...a91" }
    }
  }
}
Downstream caller (csk_)
json
{
  "mcpServers": {
    "writ-target": {
      "url": "https://api.usewrit.app/mcp/your-workspace",
      "headers": { "Authorization": "Bearer csk_3a90...f4d" }
    }
  }
}
>_

Per-client setup. Step-by-step config for each client lives on the per-client pages: Claude Desktop, Claude Code, Cursor, Windsurf.

auth ▸ logged-in web 05
06
authenticated

Your agent acts on your own Target account

Target product and price pages are public; Writ renders them in real Chromium to capture per-store availability, and can sign into your account for cart actions.

PersonasTOTP & mailbox OTP Scoped wt_ / csk_ keysStructured output

Use Writ on your own Target account and on public listings you have the right to access. Respect Target's Terms and purchase limits.

faq ▸ questions answered 07
08
faq

Target MCP questions

Is there an official Target MCP server?
Target does not publish an MCP server. Writ lets you turn the Target workflows you build (recorded, described, or discovered) into MCP tools your agent can call. They take real actions, not just read page text.
Are there ready-made Target MCP tools I can install?
Record the Target task once in the recorder, or describe it to an AI session, then publish it. Writ exposes it as both an MCP tool and a REST endpoint, usually in minutes.
How is this different from a read-only website-to-MCP tool?
Most website-to-MCP tools are readability crawlers that return page text. Writ tools act and transact: they log in, fill forms, click, extract, and complete multi-step Target tasks, then return structured output your agent can use.
Does it work with Target login and 2FA?
Target product and price pages are public; Writ renders them in real Chromium to capture per-store availability, and can sign into your account for cart actions. Because each MCP tool is a real browser workflow, it runs inside your own authenticated Target session, with TOTP and mailbox-OTP two-factor handled by Personas.
Which MCP clients can use these Target tools?
Any MCP-compatible client: Claude Desktop, Claude Code, Cursor, Windsurf, and others. Point the client at your Writ MCP URL and authenticate with a wt_ key (or a csk_ consumer key for downstream callers).
Is it allowed to use Target this way?
Use Writ on your own Target account and on public listings you have the right to access. Respect Target's Terms and purchase limits. Writ runs on your own accounts, with your own credentials and data, on sites you are authorized to use.
rec ▸ ready

Give your agent Target tools

Build a workflow, publish it as an MCP tool, and call it from any MCP client. Get started free.