Designs, generates and packages an MCP server or an agent skill, with the current SDK documentation at hand.
Built for: The developer wrapping their product's API for Claude and other clients, who wants it to work the first time a stranger installs it.
The button opens the checkout, where the 3 servers and 6 skills are listed one by one. Every member is free, so the checkout lists them and asks for no card.
The reason to buy a stack rather than five listings: each line below needs more than one member connected at the same time.
Check the official MCP Registry for existing servers before starting, then generate a FastMCP skeleton from a description.
Design the tool names, descriptions and schemas so a model picks the right tool, using the current SDK documentation.
Add an interactive UI to a tool, or package a local server as an MCPB bundle for one-click install.
Write the companion agent skill and test when it triggers.
What each part contributes, and why it was picked over the obvious alternative. Prices and permissions are read from the listings, so nothing here can disagree with the catalogue.
Current, version-specific documentation and code examples for libraries, including the MCP SDKs, instead of what the model remembers.
Generates a FastMCP server from a plain-English description, then validates and inspects what it generated.
Queries the official MCP Registry — servers by name, their versions and details — for checking what already exists, and which names are taken, before you build.
Anthropic's entry point for building an MCP server: transport, tools, auth and how to test it.
Anthropic's guide to designing tools that models use well, in Python with FastMCP or in TypeScript.
Adds interactive UI to an MCP server — forms, pickers, dashboards rendered in the conversation.
Packages a local MCP server as an MCPB bundle with its runtime, for one-click install.
Anthropic's skill for writing, testing and tuning agent skills, including evaluations of when they trigger.
The Claude API reference — models, tool use, streaming, caching — for the server that calls a model itself.
Every member is free, so the monthly figure is $0 whatever you do with the stack: nothing here is billed by the call, and only adding a paid member could change it.
| Part | What you are paying for | Monthly, as quoted |
|---|---|---|
| Context7 | Free | — |
| mcpforge | Free | — |
| Registry | Free | — |
| Skills | ||
| Build an MCP Server | Free · context cost only | — |
| MCP Server Development Guide | Free · context cost only | — |
| Build an MCP App (Interactive UI Widgets) | Free · context cost only | — |
| Build an MCPB (Bundled Local MCP Server) | Free · context cost only | — |
| Skill Creator | Free · context cost only | — |
| Building LLM-Powered Applications with Claude | Free · context cost only | — |
| Everything above | $0 of servers plus $0 of skills, the same in a quiet month and a busy one | $0/mo |
Subscriptions at their monthly plan price; one-off licences spread over 12 months.
Install only these and the bill is nothing: 3 servers and 6 skills, and all 15 tools.
Nothing in this stack carries a call allowance, so nothing here can run out before the month does. The bill is decided when you install it, not when you use it.
The spend cap is one figure for the whole account, set in your dashboard and checked when an order is placed. On the call path what stops a retry loop is the plan's call allowance and its rate limit, and a refused call is never billed.
The steps below come before the install lines: what to decide and scope before connecting anything, in the order it matters. The lines themselves follow, part by part, for the client you pick.
npx mcprush@latest stack add mcp-builder writes all 3 servers into Claude Code’s config in one go — or run the lines below one by one. The 6 skills are not included — add them with lines 04–09.
ANTHROPIC_API_KEY — Anthropic API key that generate, update and plan call the model with, billed to that accountSet it where you start Claude Code — export ANTHROPIC_API_KEY=… in your shell profile on macOS and Linux, setx ANTHROPIC_API_KEY "…" on Windows, then open a new terminal. Claude Code reads it at start; the config keeps only the name.
# 01 · Context7
claude mcp add --scope user --transport http context7-mcp https://mcp.context7.com/mcp
# 02 · mcpforge
claude mcp add --scope user --env 'ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY}' --transport stdio mcpforge '--' uvx fastmcp-builder
# 03 · Registry
claude mcp add --scope user wei-registry-mcp '--' npx -y mcp-registry-mcp-server
# 04 · Build an MCP Server
npx mcprush@latest skill add anthropics/build-an-mcp-server
# 05 · MCP Server Development Guide
npx mcprush@latest skill add anthropics/mcp-server-development-guide
# 06 · Build an MCP App (Interactive UI Widgets)
npx mcprush@latest skill add anthropics/build-an-mcp-app-interactive-ui-widgets
# 07 · Build an MCPB (Bundled Local MCP Server)
npx mcprush@latest skill add anthropics/build-an-mcpb-bundled-local-mcp-server
# 08 · Skill Creator
npx mcprush@latest skill add anthropics/skill-creator
# 09 · Building LLM-Powered Applications with Claude
npx mcprush@latest skill add anthropics/building-llm-powered-applications-with-claudenpx).uvx, docs.astral.sh/uv).Set-ExecutionPolicy -Scope CurrentUser RemoteSigned once, or type npx.cmd instead of npx (codex.cmd, gemini.cmd, grok.cmd likewise). PowerShell 7 needs neither.Nothing is connected until you approve it in Claude Code. mcprush is not in the path and meters nothing here.
Installed together, these tool surfaces add up. It is the first thing a security reviewer asks for, so what has been counted — and what nobody has counted yet — is on the page rather than in a PDF.
| Member | Share of the write surface | Write tools |
|---|---|---|
| mcpforge | 1 |
A stack's blast radius is the union of its members, not the worst of them. The 1 write-capable tools are the ones worth an argument in a review; the 14 read-only ones are worth a glance at what they can read.
A stack is a default, not a verdict. These are the substitutions the maintainer would make, and what each one costs or saves.
The maintainer lists no alternatives for this stack.
Written by the maintainer, kept on the page rather than in a support thread.
It does not publish anything. Registry entries, npm and PyPI releases are made from your own accounts.
Generated code is a starting point. Review it, test it and scan it before anybody installs it.
Context7 is a hosted documentation service; what it returns is only as current as its index.
A stack has an owner: whoever keeps it re-checks the combination when a member changes, and the members themselves are published by the people named on each row.
Tools to inspect code changes, run checks and prepare a pull request review.
Tools to investigate a bug report and work toward a reproducible test case.
Tools to investigate alerts, check recent changes and follow the runbook.