Free AI & MCP Tools for Builders
Everything you need to build, secure and work with AI — in one place, running in your browser. If you're wiring up AI agents and MCP servers, there's a full toolkit here: generate and validate MCP configs, build tool/function-calling schemas for Claude, OpenAI and MCP, inspect a tools/list, lint a system prompt, redact secrets before you paste them, and scan servers and untrusted text for prompt-injection and tool poisoning.
For everyday AI writing and coding, the bring-your-own-key helpers connect straight from your browser to OpenAI, Groq or any OpenAI-compatible provider — summarize long text, rewrite and change tone, and explain code. Your API key lives only in your browser and requests go directly to the provider, so neither your key nor your text ever passes through our servers.
Also useful for ai
These live under Developer, but they come up just as often here.
MCP Security Scanner
Scan an MCP server config for prompt-injection, exfiltration & tool poisoning.
MCP Config Generator
Generate a valid MCP server config for Claude, Cursor, VS Code & Windsurf.
MCP Config Validator
Validate an mcp.json / claude_desktop_config.json and catch common mistakes.
Prompt Secret Redactor
Mask API keys, tokens & PII in a prompt before you send it to an LLM.
Prompt Injection Checker
Scan untrusted text for prompt-injection patterns before feeding it to an agent.
Tool Definition Generator
Generate a tool/function-calling schema for MCP, Anthropic (Claude) & OpenAI.
MCP tools/list Inspector
Inspect & validate an MCP tools/list response — params, required fields & schema issues.
System Prompt Linter
Review a system prompt for role clarity, guardrails & prompt-injection resistance.
Tool Definition Converter
Convert a tool/function schema between MCP, Anthropic (Claude) & OpenAI formats.
MCP Config Merger
Merge multiple MCP server configs into one file — deduped, with conflicts flagged.
Prompt Template Manager
Fill {{variables}} in a reusable prompt with a live preview and missing-value checks.
WebMCP Tool Generator
Generate WebMCP tool definitions — declarative HTML & imperative JS — to make your site agent-ready.
WebMCP Readiness Checker
Check your page HTML for WebMCP tools and get agent-readiness fixes — is your site WebMCP-ready?
llms.txt Generator
Build an llms.txt so AI assistants can find and understand your key pages.
AI Crawler Blocker (robots.txt)
Block ChatGPT, Claude & other AI bots from your site — without hurting SEO.
Build & secure AI agents and MCP servers
- MCP Security Scanner — flag prompt-injection, tool poisoning and data-exfiltration in an MCP server config.
- MCP Config Generator & Validator — create and lint mcp.json / claude_desktop_config.json for Claude, Cursor, VS Code and Windsurf.
- Tool Definition Generator — one tool description exported as MCP, Anthropic and OpenAI function-calling schemas.
- MCP tools/list Inspector — read and validate a server's tools, parameters and required fields.
- System Prompt Linter — grade a prompt for role clarity, guardrails and prompt-injection resistance.
- Prompt Secret Redactor & Prompt Injection Checker — mask secrets before you paste, and screen untrusted text before an agent reads it.
AI writing & coding helpers (bring your own key)
- AI Text Summarizer — condense articles, reports and transcripts into a TL;DR, paragraph or bullets.
- AI Paraphraser / Rewriter — reword text and shift tone between professional, casual, formal or concise.
- AI Code Explainer — step-by-step explanations of a snippet, plus likely bugs to watch for.
Why these run in your browser
The MCP and agent tools are pure client-side utilities — the config, schema, prompt or text you paste is analysed on your own device and never uploaded, which matters when that text may contain API keys or internal prompts. The writing helpers use bring-your-own-key: instead of capping usage, adding a margin, or routing your text through our infrastructure, they send it straight to the provider you choose at the price your provider charges. Providers such as Groq offer a free tier, so this can cost nothing at all.
Frequently asked questions
What are the MCP and agent tools for?
They help you set up and secure AI agents that use the Model Context Protocol (MCP). You can generate and validate server configs, build tool schemas for Claude/OpenAI/MCP, lint system prompts, and scan servers and untrusted text for prompt-injection and tool poisoning — all in your browser.
Do the MCP tools upload my config or prompts?
No. The MCP, schema, prompt and redaction tools run entirely in your browser, so configs and prompts — which can contain secrets — stay on your device.
What does bring-your-own-key mean for the writing tools?
You supply an API key from a provider like OpenAI or Groq. Requests go directly from your browser to that provider, so you control the cost and we never see your data or your key.
Can I use these for free?
The MCP and agent tools are completely free with no key needed. For the writing helpers, several providers — Groq among them — offer a free API tier that's enough for regular use.
Which providers work with the writing tools?
Any OpenAI-compatible endpoint, including OpenAI, Groq and OpenRouter, plus custom endpoints you configure yourself.