Branderux

com.branderuxPublished by branderux.comOfficial publisherv0.1.3Updated Oct 26, 2026

One message in, a full agentic application out: website and MCP app, live. Built from any AI client.

VerifiedStreamable HTTPWeb executableDeveloper ToolsBusiness & CommerceAI & ML

Overview

AI-generated overview

Official BranderUX MCP server that lets an AI client build, configure and publish agentic web and MCP applications from a business description.

What you can do
  • Read BranderUX docs, integration snippets and reference templates
  • Create and manage projects, brand settings and screens
  • Write, validate, publish and preview custom TSX elements
  • Manage API keys, hosted agent config, skills and connectors
  • Define business entities and records, and query leads, orders, conversations and visitor stats
  • Set home and fixed screens, verify bindings and publish the site as a website plus an MCP app
When to use it
Use it when you want an assistant to generate and operate a full BranderUX application — site, MCP app, hosted agent and business data — from a short business description, or to work the leads and conversations that app collects.
Requirements
Remote streamable HTTP endpoint at mcp.branderux.com; a BranderUX account, signed in through browser OAuth 2.1 with PKCE on the first tool call. No API keys or environment variables are needed for the hosted server.
Before you install
Tools create, update and delete projects, screens, elements, API keys, records and skills, and publish live sites; destructive tools require an explicit confirm flag. API key management and connector credentials are sensitive. Owner-data tools read customer leads, conversations and visitor data, and CRM sending moves customer data to third parties.

Installation

In SourceWeft

  1. Open Branderux in the dashboard and add it to a workspace.
  2. Connect the required credentials. They are stored privately and never shown on public pages.
  3. Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Use with an agent

Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.

Read https://sourceweft.com/mcp/com.branderux%2Fmcp/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.

Read the install guide the agent follows

Other MCP clients

Add this to your client's mcpServers config. The client completes OAuth sign-in on first connect.

{
  "mcpServers": {
    "branderux": {
      "type": "http",
      "url": "https://mcp.branderux.com/mcp"
    }
  }
}

README

BranderUX MCP

BranderUX builds agentic applications, in minutes: describe a business and a real agent builds its full application, published as a website and an MCP app, with every screen generated live, per customer, per question. This server is how any MCP client (Claude, ChatGPT, Cursor, VS Code) builds and controls those applications.

This is the official BranderUX MCP server. It gives an AI agent real control over BranderUX projects, brand, custom elements, screens, API keys, plus the verified reference docs it needs to integrate the SDK correctly.

Connect: https://mcp.branderux.com/mcp · Docs: https://branderux.com/mcp

bash
claude mcp add --transport http branderux https://mcp.branderux.com/mcp

No API keys: the first tool call opens your browser for a one-click BranderUX sign-in (OAuth 2.1 + PKCE, scoped and revocable). Anyone with a BranderUX account connects. See what it builds: nova.branderux.app and the five live examples on branderux.com.

Tools

Knowledge (no scopes needed, signing in is still required to reach the server): get_started · read_doc · search_docs · get_integration_snippet · list_templates · get_template. The widget snippet of get_integration_snippet is the one-line script tag for a plain website (Wix, WordPress, Shopify, static HTML), carrying data-preload="eager" and data-conversions="on" (new leads, bookings, orders and requests are reported to the tags already on the site). list_templates and get_template read the five reference builds as worked examples to adapt, never to copy.

Projects (projects:*), for the signed-in user (projects they own or manage): whoami · list_projects · get_project · create_project · update_brand_settings · update_project_settings · delete_project

Screens (projects:write): list_screens · get_screen · put_screen · delete_screen, reads fetch the project aggregate; writes are atomic per-screen (row-locked server merge: parallel saves of different screens are safe).

Custom elements (elements:*): list_elements · get_element · list_element_versions · create_element · publish_element_version · preview_element · delete_element, your agent writes the TSX; the server pre-flight validates it (compile + sandbox import allowlist + export contract) before publishing. In clients that support MCP Apps, create_element, publish_element_version and preview_element render the element LIVE in the panel, demo props applied, clicks showing the exact query they would send.

API keys (keys:manage): create_api_key · list_api_keys · set_key_origins · revoke_api_key

Hosted agent: upsert_agent_config · get_agent_config, give the project its own agent: persona, policies (language and clock, login, access, handoff, write consent), answer quality and budgets.

Business data: define_entity · list_entities · seed_records · update_record · list_entity_records, managed records or live store/API feeds; every entity becomes a query tool for the agent (add-only writes by default).

Owner data (projects:read to read, projects:write to change): query_records · aggregate_records · get_record · list_conversations · aggregate_conversations · get_conversation · list_visitors · get_stats · update_record_workflow · correct_record · add_record · record_opt_out · delete_record · reply_links · get_crm_status · send_record_to_crm, the owner's own AI reads the leads, bookings, orders and requests the hosted agent collects, with its conversations, signed-in visitors and numbers, and works them one record at a time: every change is logged on the record (a delete erases the record with its log), visitor text comes back marked as data, nothing is sent to a customer (reply_links prepares a draft the owner sends), and where customer data goes stays the owner's or a manager's choice in the app. The review-my-week prompt goes through the owner's week or month with these tools. The docs: read_doc owner-data.

Connectors: set_connector_credential · probe_api, vaulted credentials and a probe so tools are wired against the real API shape.

Skills: upsert_skill · list_skills · delete_skill, SKILL.md behavior packs that ride every conversation.

Home and fixed screens: set_home_screen · set_fixed_screens · list_fixed_screens · verify_canned_screens, a designed first paint and designed answers to the business's recurring questions, each replayed with live data bindings at zero model cost per visit; verify_canned_screens runs every binding the way serving does, so a screen is proven to replay before publish_site.

Publish: publish_site · get_site, one call and the full application is live: a website for people, an MCP app for AI agents.

Playground (no project needed): generate_screen, renders a real branded, interactive screen in the panel with demo data, powered by the same published @brander/mcp-tools package customers install. Ask for a storefront, analytics or order-flow screen to see actual BranderUX output before building anything.

Every destructive tool requires an explicit confirm: true.

Local development

bash
npm installBRANDER_API_BASE=http://localhost:8080/api/v1 npm run dev   # http://localhost:3010/mcp
EnvDefaultPurpose
BRANDER_API_BASEhttp://localhost:8080/api/v1BranderUX API base
BRANDER_APP_BASEfrom BRANDER_API_BASE: https://dev.branderux.com for api-dev, https://branderux.com for api, else http://localhost:3000The BranderUX web app: playground links, canned-screen verification and query_records on a live catalog
MCP_RESOURCE_URLhttp://localhost:3010Public URL of this server (OAuth resource id)
OAUTH_ISSUER_URL= BRANDER_API_BASEAuthorization server issuer

The server is stateless: one MCP server instance per request, bound to the caller's bearer. Deploy target is Vercel (api/mcp.ts + api/oauth-protected-resource.ts).

License

MIT

Source: README.md at commit e58c9df

Tools

0
This server asks you to sign in, so SourceWeft could not read its tools without an account and has not indexed them. What a connected account sees can depend on that account and its permissions.

Version history

1
  1. v0.1.3LatestSep 16, 2026

Score

85% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 daysRequired25 pts
  • Auth declaredRequired10 pts
  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts