SAY WHATYOU MEAN.

We’ll tell the AI what you meant.

Write the messy, half-formed instruction already in your head. SayWhat turns it into a clear, context-aware prompt your coding agent can actually work with.

$4.98

Cursor first. Other agents next.

  • Cursor
  • Claude Code
  • Codex
  • Windsurf
  • VS Code
  • Copilot
  • Cline
  • Zed
  • JetBrains
  • Replit
  • TRAE
Composer

Sample workflow

A real messy thought. Watch it compile.

Before / after

MESSY IN.
CLEAR OUT.

Before

“make login work again”

After

Debug the existing login flow. Reproduce the failure, inspect the current auth and session path, and fix the root cause. Do not rebuild authentication. Verify sign-in, sign-out, and a protected page.

More than grammar

SAYWHAT UNDERSTANDS
MORE THAN SPELLCHECK.

  1. 01

    Intent

    What you actually want done.

  2. 02

    Scope

    Where the work starts and stops.

  3. 03

    Context

    What already exists in the project.

  4. 04

    Constraints

    What not to break, invent, or expose.

  5. 05

    Acceptance

    How we’ll know it’s finished.

  6. 06

    Testing

    What to verify before calling it done.

  7. 07

    Ambiguity

    The fuzzy words that hide decisions.

  8. 08

    Risk

    Auth, data, deploy, secrets.

Modes

NOT LONGER.
BETTER.

SayWhat isn’t designed to turn every sentence into a 500-word prompt. It decides what the task actually needs.

LIGHT

“the settings page is messy on mobile”

Fix the settings page layout on small screens. Keep the current information architecture and components; tighten spacing and wrapping so actions stay usable. Check a phone-width viewport.

AUTO

“rename UserCard to ProfileCard and update imports”

Rename UserCard to ProfileCard and update all imports. Don’t change behavior or styling.

FULL

“add another pricing tier”

Add a third pricing tier using the existing pricing configuration and checkout flow. Do not create a second payment path. Inspect the current plan model, Stripe (or existing provider) products, and pricing UI first. Match copy, limits, and upgrade behavior to the current tiers. Test checkout for the new tier and a switch from an existing plan.

First integration

BUILT FOR
CODING AGENTS.

  • CursorAvailable — early access
  • Claude CodePlanned
  • CodexExploring
  • WindsurfPlanned
  • VS Code AgentsExploring
  • MoreExploring

How it compiles

THE INTENT LAYER.

You

“make checkout better”

SayWhat
  • Intent
  • Context
  • Scope
  • Constraints
  • Acceptance
  • Testing
AI agent

“Understood.”

Opt-in metadata

REPO-AWARE.
NOT NOSY.

The Cursor panel can notice Next.js, Cloudflare, Stripe, Drizzle, Tailwind, Vitest — if you turn that on. It still does not upload your source.

In the Cursor panel

  • Next.js
  • Cloudflare
  • Stripe
  • Drizzle
  • Tailwind
  • Vitest

In the Cursor panel

PROMPT QUALITY.

A heuristic score after you enhance — not a public grade, not a gimmick. Shown so you can see what got clearer.

Prompt Quality

62 → 94

  • Objective✓
  • Context✓
  • Scope✓
  • Constraints✓
  • Acceptance✓
  • Testing✓
  • Ambiguity✓

WHO THIS IS FOR.

You shouldn’t need a prompt-engineering degree to change a button.

  • Founders

    Ship the thing without writing a spec novel.

  • Designers

    Describe the feeling. Get a change the agent can execute.

  • PMs

    Capture the request once. Don’t babysit the wording.

  • Vibe coders

    You shouldn’t need a prompt-engineering degree to change a button.

  • Juniors

    Ask for help like a human. Send it like an engineer.

  • Seniors

    Skip writing miniature specifications all day.

FAQ.

It compiles a messy instruction into a clearer, agent-ready prompt. Intent, scope, constraints, and testing get filled in when they are missing — and left alone when they are not.

saywhat.foo

SAY IT YOUR WAY.

We’ll make it agent-ready.

Where do you write prompts? (optional)

Join Early Access for the Cursor install. Enhancement is live in the panel — you still review before anything reaches your agent.