<img src="assets/cover-rounded.png" alt="moo.md - Stay present with AI" width="100%">

hope @ moo.md

Why does this exist? Why introduce theatrical friction?

Because YOU the human -> end up being the world model. The LLM powered agentic coder is just your hands with tiny muscle memory.

---

Before moo:

"Build a settings page" → starts coding → wrong pattern → rebuilds → ships something nobody asked for

After moo:

"Build a settings page" → clarifies what matters → shapes the approach → builds with confidence

---

Try it

/plugin marketplace add saadshahd/moo.md
/plugin install hope@moo.md

You say: /hope:full build a settings page and moo:

• clarifies intent — What settings? Who sees them? What should this NOT do?
• shapes design — 2 approaches with tradeoffs. You pick before writing code.
• consults experts — Surfaces what you'd miss. One actionable next step.

<details> <summary><strong>See it in action</strong> — real session, messy scenario</summary> <br>

Intent — "Make the homepage better."

[INTENT] Specific questions, not "what do you want?"

moo asks:

1. What's the main thing the homepage should accomplish?
2. What specifically feels wrong or weak about it right now?
3. What should the homepage NOT do or become?

You answer:

1. "Get more signups. We're at 2% conversion, want 5%."
2. "People don't understand what we do. Hero is too clever. Signup button is below the fold."
3. "No long-scroll landing page. No chatbot widget."

Two more rounds — product details, tech stack, traffic split, social proof assets. Each question targets a specific gap.

[ECHO] 33 words. You confirm or edit before anything proceeds.

moo echoes back: Rewrite homepage copy and layout (Next.js/Tailwind) to clearly communicate the AI email marketing value prop, move signup above the fold, and add social proof — targeting 5% conversion, especially on mobile.

[ASSUMPTIONS] Surfaces decisions you never stated — forces you to choose.

Page length — you said no long-scroll. How many sections?

• A) Max 4   B) Max 5 (you pick)   C) Max 3

Signup mechanism — what fields?

• A) Email-only + "Get Started Free" (you pick)   B) Email + company name   C) Separate signup page

Mobile target — mobile is at 1.2%. Separate target?

• A) Same 5%   B) Mobile 3.5%, desktop 5%+ (you pick)   C) No separate target

---

Shape — "How should we build this?"

[CONSULT] Experts debate. Real tension, not rubber-stamping.

moo identifies the key architectural decision: should the hero section lead with the product or the outcome?

Krug (usability): "Don't make me think. Lead with what the product does — users scan, they don't read. Clever taglines fail."

Cialdini (persuasion): "Lead with the outcome — social proof and results pull harder than feature descriptions. Show the 50M emails number above the fold."

Tension: Krug wants clarity-first ("AI email marketing for e-commerce"). Cialdini wants proof-first ("50M+ emails sent for brands like yours"). Different advice, real tradeoff.

[SHAPE] Two concrete approaches with tradeoffs. You pick before code.

	Approach A: Clarity-first	Approach B: Proof-first
Hero	"AI-powered email campaigns for e-commerce" + signup	"50M+ emails sent" + brand logos + signup
Strength	Instant comprehension	Instant credibility
Risk	Feels generic without differentiation	Confusing if visitor doesn't know the category
Best when	Product category is unfamiliar	Product category is obvious, trust is the gap

You pick Approach A — visitors don't know the category yet. Proof moves to section 2.

---

Brief — precise spec, no code written yet

Objective	Rewrite homepage copy and layout. 5% desktop / 3.5% mobile conversion.
Non-goals	No visual redesign, no long-scroll, no chatbot, no CMS changes.
Acceptance	Hero = "result + how." Signup in first viewport. Social proof within 2 scrolls. Single CTA style.
Stop conditions	Conversion drops below 2%. Lighthouse below 90. Page exceeds 5 sections.

From "make the homepage better" to a spec with stop conditions — before a line of code.

</details>

---

Gratitude

Nate B. Jones · Superpowers · Farnam Street