AI Death

Will a markdown file replace your company? Type a sentence, find out.

Live: https://ai-death.vercel.app

A satirical web app and CLI that rates a company description on its likelihood of being replaced by a Claude Skill — a .md file. Five tiers from IMMORTAL to DEAD, deterministic in-browser scorer, optional context controls, an animated court HUD that does way too much.

Try it in 5 seconds

npx ai-death

Boots a local server at http://127.0.0.1:5173. No install, no API key, no account.

What you get

Click any row to see the live verdict on ai-death.vercel.app:

Description	Tier	Score
"AI assistant that summarizes meetings and writes follow-up emails using prompt templates."	DEAD	94
"Two-sided marketplace connecting freelance plumbers with homeowners, with payments and dispute resolution."	FORTRESS	26
"HIPAA-compliant electronic health records system used by 200 hospitals. Integrates with lab equipment, insurance billing, and physical pharmacy dispensing."	IMMORTAL	0

The court delivers a verdict, a death probability, three sharpened roast notes, and a mock .md skill that would replace the product.

DEAD — wrappers, summarizers, drafters, "AI-powered X"	FORTRESS — marketplaces, networks, deep integrations

IMMORTAL — compliance gates, regulated buyers, atoms, contracts

Why this exists

A lot of SaaS today is a UI wrapper on logic that an LLM can replicate. A subset isn't, because it has real moats: regulators, atoms, networks, contracts. This is a court that helps you tell the difference, in one sentence.

The premise is mean enough to be useful. If you can't say what your product does in a way that survives the court, the product probably doesn't survive the next year of model releases either.

Tier ladder

Tier	Score	Vibe
IMMORTAL	0 – 19	Compliance, hardware, real-world ops. Stay of execution granted.
FORTRESS	20 – 39	Network effects, marketplaces, deep integrations.
SWEATING	40 – 59	Solid product, but the LLM is closing in.
THIN ICE	60 – 79	Mostly a UI on top of GPT/Claude with a logo.
DEAD	80 – 100	Could be a one-page markdown skill. Funeral arrangements.

Optional context

Below the description, three pill rows can sharpen the verdict by ±42 points and pull tier-aware roast notes:

• Stage — Idea / Pre-launch / Has users / Paying customers / Enterprise contracts
• Buyer — Consumer / SMB / Mid-market / Enterprise / Regulated
• Surface — Pure software / Hardware / Physical ops / Compliance gates

Same description, very different verdict. An idea-stage consumer wrapper reads THIN ICE / DEAD. The same idea pivoted to enterprise-contract regulated compliance gates lands FORTRESS / IMMORTAL.

Shareable verdicts

Pre-bake a verdict by encoding the description (and optional context) in the URL hash:

https://ai-death.vercel.app/#desc=ENCODED_TEXT&stage=paying&buyer=enterprise&surface=compliance

The form auto-submits on load, so you can drop these into Slack/X/HN comments. A few primed examples:

• DEAD — chatbot wrapper
• THIN ICE — note-taking app
• SWEATING — vertical SaaS for dental clinics
• FORTRESS — freelance plumber marketplace
• IMMORTAL — HIPAA EHR for hospitals

How the scorer works

100% client-side, deterministic, no API calls.

1. Start at 50.
2. Decrement for moat keywords: hipaa, hardware, marketplace, payments, compliance, network, enterprise, regulated, … (~40 patterns).
3. Increment for wrapper keywords: summariz, chatbot, wrapper, cold email, template, resume rewrite, … (~40 patterns).
4. Apply the optional Stage / Buyer / Surface shifts.
5. Length signals: very short input is vague (penalty), very long is specific (small mercy).
6. Add deterministic noise from a hash of the input + context, so identical inputs are stable but the spread feels alive.
7. Clamp 0 – 100, map to tier.

Lexicons live at the top of app.js: MOATS, REPLACEABLE, TIERS, TAGGED_ROASTS, CONTEXT_SHIFTS, CONTEXT_ROASTS, plus the per-tag verb map inside buildSkill. Tweaking is one regex away.

Repo

.
├── index.html           ← markup + verdict-card template
├── styles.css           ← design tokens, tier colors, court-HUD animations
├── app.js               ← scorer, roast bank, render
├── bin/ai-death.js      ← zero-dep Node http server (Node 18+)
├── package.json         ← npm bin, files allowlist
├── DESIGN.md            ← design system source of truth
└── docs/screenshots/    ← README assets + 1200×630 OG card

Run from source

It's static HTML/CSS/JS. Any static server works.

git clone https://github.com/vnmoorthy/ai-death.git
cd ai-death
python3 -m http.server 5173

Then open http://localhost:5173.

Deploy

Drop the folder on any static host:

• Netlify / Vercel / Cloudflare Pages — drag the folder into the dashboard.
• GitHub Pages — Settings → Pages → deploy from main / /.
• Anywhere with python3 -m http.server — same as local.

Zero environment variables. Zero keys.

What this is NOT

• Not a real evaluation tool. Don't make decisions based on the verdict.
• Not a lead-gen page. No email capture, no waitlist, no analytics.
• Not advice. Probably not even useful. Definitely funny if your description is bad.

License

MIT © vnmoorthy