Backend Unslop
Audit API and database code for validation, query, error-handling, and security mistakes.
Connect your AI assistant to the tools and data it needs. Each listing shows what the server does, what it needs, and the config to copy.
Audit API and database code for validation, query, error-handling, and security mistakes.
Review English and Bahasa Indonesia product copy for vague wording, readability, and length.
Review AI-written React, Next.js, and Tailwind code for implementation mistakes.
Review native and cross-platform mobile code for crash risks, permissions, and accessibility gaps.
Explore databases in a desktop workbench with an AI SQL assistant and a local read-only MCP mode.
Find ineffective tests, run the available suite, and report whether tests actually executed.
Check interface contrast, accessibility, design tokens, and interaction states.
MCP stands for Model Context Protocol. It is an open standard that lets an AI assistant call outside tools, like querying a database or running a check on your code. An MCP server is the program that offers those tools.
You add a server to your assistant's MCP settings. Claude Code, Claude Desktop, Cursor, and Codex each read a short config that names the server and how to start it. After a restart, the assistant can call the server's tools while it works.
All three support MCP. The config file and keys differ a little between apps, so check each project's setup notes for the exact format.
No. You run each server on your own machine or account. This site lists them and links to their source.
No. An MCP server gives your assistant tools to call. A skill gives it written instructions for a task. Some projects ship both, and those listings appear under both types.