diy-mcp-connector

diy-mcp-connector is a skill for Claude Code from sshh12/claude-plugins. It costs 66 tokens per session (3,124 once invoked), scanned A, original, MIT.

A guide that builds a dedicated MCP server for one web app. An MCP server is a small program that lets Claude and other compatible tools call an app's functions through defined tools.

In plain words
What is it for?
Use it to connect a web app to Claude Code or Claude Desktop, investigate its API, design and implement tool calls, handle authentication and responses, test the server, and deploy it.
Why use it?
It turns a web app's API into focused Claude tools and covers discovery, authentication, security, testing, and deployment in a defined sequence.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: mentions subagents; mentions Claude Code; built for openclaw.

Part of the diy-mcp-connector plugin — 1 skill shipped together

Good fit Use it to connect a web app to Claude Code or Claude Desktop, investigate its API, design and implement tool calls, handle authentication and responses, test the server, and deploy it.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/sshh12/claude-plugins/diy-mcp-connector
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add sshh12/claude-plugins --skill diy-mcp-connector
Clone the repo
git clone --depth 1 https://github.com/sshh12/claude-plugins

Made for: Claude Code.

Or install diy-mcp-connector, the plugin that ships this one along with the rest of its 1 skill.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for diy-mcp-connector

README.md
[![agentmods](https://agentmods.dev/badge/skills/sshh12/claude-plugins/diy-mcp-connector/github.svg)](https://agentmods.dev/skills/sshh12/claude-plugins/diy-mcp-connector)
Your own site
<a href="https://agentmods.dev/skills/sshh12/claude-plugins/diy-mcp-connector"><img src="https://agentmods.dev/badge/skills/sshh12/claude-plugins/diy-mcp-connector/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for diy-mcp-connector

Your own site · 80×15
<a href="https://agentmods.dev/skills/sshh12/claude-plugins/diy-mcp-connector"><img src="https://agentmods.dev/badge/skills/sshh12/claude-plugins/diy-mcp-connector.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 66 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,124 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00066 $0.03124
Opus 5 $0.00033 $0.01562
Sonnet 5 $0.00013 $0.00625
Haiku 4.5 $0.00007 $0.00312

Measured 10d ago against content hash 5a981f5e2762, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

diy-mcp-connector scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.

The scan reads SKILL.md. This mod also ships 7 executable files (scripts/analyze-har.py, scripts/auth.js, scripts/csrf.js, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

plugins/diy-mcp-connector/skills/diy-mcp-connector/SKILL.md · 248 lines

How it starts

The opening of the file, as written. The whole thing — 248 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Overview

This skill runs in Claude Code and builds a standalone MCP server that turns one web app's API into purpose-built tools. The connector you build works with Claude Code, Claude Desktop (Cowork), and any other MCP-supporting application. The result is a Node.js server that handles authentication, response formatting, and tool routing — one dedicated server per app.

All 9 stages are mandatory. Complete each in order — do not skip ahead. Each stage has a gate condition that must be met before proceeding.

Stages can loop. Auth or API issues discovered in Stages 6-7 may require returning to Stage 2 to reclassify the app type, then re-running Stages 5-7. This is expected — the linear numbering is the happy path, not a one-way gate.

Prerequisites

  • Node.js 18+ and npm
  • Python 3 (for HAR analysis script)
  • Google Chrome (for auto-login via CDP)
  • @modelcontextprotocol/sdk npm package (installed during build)
  • Optional (for live exploration): Claude for Chrome, brw, Playwright MCP, or Chrome DevTools MCP — any browser MCP server enables live API discovery instead of manual HAR capture

Getting Started

  1. Ask the user which web app they want to connect
  2. Detect available browser tools (claude mcp list) — see references/1-capture-api.md for the detection table and preference order
  3. If no browser tools found, ask if they have HAR files or want to set up a browser tool (Claude for Chrome is the easiest)
  4. Create the project directory: <app>/
  5. Copy the analyze script: cp "${SKILL_DIR}/scripts/analyze-har.py" <app>/scripts/

Stage 1: Capture API Surface

See references/1-capture-api.md

  • If api-key-auth with documented API: skip HAR, document endpoints from docs/existing code
  • Otherwise: detect available browser tools via claude mcp list
  • Choose capture method: HAR files, live exploration (using detected tool), or both
  • Capture API traffic across all major app sections (not just one page)
  • Save HAR to <app>/har/<app>.har (if using HAR method)

Read the full file on GitHub · 248 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 10d ago First seen · 248 lines · 66 tokens per session scan A 5a981f5e2762

Subscribe to this mod's changes

diy-mcp-connector is a skill published in the GitHub repository sshh12/claude-plugins (22 stars, last pushed 2mo ago), licensed MIT. It adds 66 tokens to every session and 3,124 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

frontmcp-development

Use when building any FrontMCP server component other than a tool (for tools, use create-tool). Covers @Resource static resources and parameterized URI templates; @Prompt reusable prompts (RAG, multi-turn); @Provider singleton dependency-injection providers (database pools, API clients); @Agent autonomous LLM agents…

agentfront/frontmcp · 196 tokens

pipefy-api-fallback

Use this skill when an MCP tool fails AND the introspection skill could not resolve the problem. This is the last-resort fallback (Tier 3): call the Pipefy GraphQL API directly using curl or httpx, authenticating with the Service Account (OAuth2) or a Personal Access Token (PAT) available as env var. Follow the 3-tier…

pipefy/ai-toolkit · 86 tokens

pipefy-introspection

Use this skill when you need to discover GraphQL type shapes, mutation signatures, enum values, or execute arbitrary GraphQL as a fallback. This is the first fallback tier (Tier 2) when dedicated MCP tools fail or don't exist for an operation. 7 MCP tools.

pipefy/ai-toolkit · 61 tokens

pipefy-ipaas

Use when the user wants to build, test, publish, or manage iPaaS (Advanced Automations) flows: multi-step integrations with external apps (Slack, Gmail, Google Sheets), incoming webhooks, schedules, routers, code steps, or iPaaS data tables. MCP-only, driven through 4 meta-tools over a per-pipe catalog discovered at…

pipefy/ai-toolkit · 104 tokens

pipefy-members-email-webhooks

Use this skill when the user wants to manage pipe membership, send or read card inbox emails, use email templates, or manage webhooks. Covers 12 MCP tools.

pipefy/ai-toolkit · 41 tokens

api-designer

Design RESTful APIs with best practices for consistency and usability.

athola/skrills · 15 tokens