Prompt guides

Plain-language guides to writing AI prompts that work, and tools that write and improve prompts for you.

• Improve a prompt that keeps missing the mark Paste a prompt that gets generic or wrong answers, say what went wrong, and get back a clearer, complete prompt. Free to start, no account needed.
• Coding prompts your AI assistant can build from Turn a feature idea into a brief your AI coding assistant can build and verify: the outcome, existing code, limits, tests and what done means. Free to start.
• Looking for a PromptPerfect alternative? PromptPerfect is no longer available after September 1, 2026. LoopTypes writes and improves text prompts for the AI tools you already use. Free to start.
• How to write a good AI prompt A good prompt tells the AI six things: the goal, the context, the must-haves, the limits, the format and how to check its work. Each one, with examples.
• What prompt made this answer? You cannot recover the exact prompt behind an AI answer, but you can write one that produces answers like it. Here is how, by hand or in one step.
• How to write a CLAUDE.md file What a CLAUDE.md is, where it goes, what to put in it and how long to make it, with an example you can copy and a tool that writes one from your README.
• CLAUDE.md examples from real projects 10 real CLAUDE.md files from open-source projects, graded out of 100 and shown whole, with what each does well and a way to make yours.
• What is AGENTS.md, and how do you write one? AGENTS.md is an open instruction file for AI coding agents. What it is, which tools read it, what to put in it, and 8 real files, graded.
• Grade your repository’s agent file and add a badge Grade the CLAUDE.md, AGENTS.md, .cursorrules or SKILL.md in a public GitHub repository out of 100, and add a grade badge to your README.
• How to prompt ChatGPT, Claude and Gemini What OpenAI, Anthropic and Google each recommend for prompting their models: where the advice agrees, where it differs, and what to change for each tool.
• Prompting FAQ Plain answers about AI prompts: what they are, why they change the answer you get, how LoopTypes writes them, and what each agent file is for.