agentleFS
scottconversescottconverse/landing-page/SKILL.md

landing-page

Skill0 starsChanged 10 days agoMITView on GitHub

Investigate an entire software repository and BUILD a real landing-page website from what the code actually contains — real source files, a design system, responsive layouts, working navigation, build scripts, and a deployment path. The website is the deliverable; the README is a short supporting artifact. This is not a README-writing task and not a design proposal: create the files, run the site, look at it, fix what's wrong, and leave the repository with a polished site ready to deploy. Use whenever the user asks for a landing page, project website, marketing site, product page, launch or beta page, docs site front page, GitHub Pages site, a "make this repo look good / look legit / presentable", a README overhaul, or a public face for a project about to be shown to users, contributors, customers, or leadership — and proactively before any repository is announced, open-sourced, or released. Also for "what does this project even do", "write the pitch for this repo", or "/landing-page".

44 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use operational-enterprise-ai, product-proof-saas and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • boraoztuncSkill

    operational-enterprise-ai

    Create or redesign enterprise AI, automation, security, and operations product pages that explain system boundaries, approvals, auditability, exceptions, and rollback. Use for dark cinematic heroes, hairline grids, metric pauses, expandable solution rows, case-study evidence, security proof, and qualified demo or waitlist handoffs.

    Also in 3 projects that use this

  • boraoztuncSkill

    product-proof-saas

    Create or redesign SaaS and AI product landing pages where a real workflow, interface, or deterministic demo is the central proof. Use for pale atmospheric shells, product UI in the hero, prompt-to-output stories, audience tabs, compact feature modules, honest pricing comparisons, and FAQ handoffs.

    Also in 3 projects that use this

  • Dragoon0xSkill

    pricing-page

    Present pricing in ways that reduce friction and increase clarity. Handles tier naming, feature comparison, anchor pricing, and plan selection psychology. Use when designing pricing pages, plan selectors, or upgrade prompts.

    Also in 3 projects that use this

  • Dragoon0xSkill

    onboarding-flow

    Design first-run experiences that create the aha moment fast. Reduces time-to-value by sequencing actions, progressive disclosure, and contextual guidance. Use when building signup flows, product tours, or empty states.

    Also in 3 projects that use this

  • shawnpangSkill

    lead-scoring

    When a founder needs to qualify inbound leads, define their ICP, build a lead scoring model, set MQL criteria, or route prospects through pipeline stages. Activate when the user mentions lead scoring, ICP, MQL, SQL, lead qualification, inbound leads, or pipeline design.

    Also in 3 projects that use this

  • claude-office-skillsSkill

    email-marketing

    Email marketing automation - campaign creation, sequence building, A/B testing, deliverability optimization, and analytics

    Also in 4 projects that use this

  • +2 more from boraoztunc

The file

We split this skill into the rules it gives an agent

10 rules. 0 are common, 10 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 39 edited versions we read keep these.

  • 10

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Create the files → run the project → look at the result → fix the visual and functional defects → iterate until it is polished.

    line 43
  2. Distinctive:

    Do not rewrite, summarize, restyle, or expand the README.

    line 55
  3. Distinctive:

    Never mistake interfaces for implementations, mocks for working integrations, tests for production features, TODOs for roadmap commitments, …

    line 65
  4. Distinctive:

    Avoid *powerful, revolutionary, cutting-edge, seamless, robust, next-generation, future-proof, enterprise-grade, all-in-one, best-in-class, …

    line 109
  5. Distinctive:

    Avoid buzzwords, clichés, hype, fake urgency, exclamation marks, generic startup language, vague promises, fake social proof, repetitive …

    line 151
  6. Distinctive:

    Prefer concrete verbs, named capabilities, explicit workflows, real constraints, actual supported platforms, actual deployment methods, …

    line 155
  7. Distinctive:

    Verify every link before it ships.

    line 163
  8. Distinctive:

    No dead placeholder links in the rendered page: When a resource doesn't exist, omit it or leave a code comment for maintainers.

    line 167
  9. Distinctive:

    Prefer real project evidence for imagery, in this order: actual product screenshots → repository diagrams → diagrams generated from the …

    line 169
  10. Distinctive:

    Write it after the site, and keep it short.

    line 179

“Common” is measured: kept by at least 6 in 10 of the 39 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Landing Page

You are an elite web designer, frontend engineer, UX writer, product strategist, software architect, technical investigator, accessibility specialist, and open-source maintainer.

Inspect an entire repository and build a complete, beautiful, production-quality landing-page website from what the repository actually contains.

The deliverable is a website with real source files — styling, responsive layouts, working navigation, polished hierarchy, build scripts, deployment support. The README is a short supporting artifact you write afterward. Not the other way around.

The result should stand credibly beside Supabase, Vercel, Astro, Tailwind CSS, Docker, Home Assistant, Linear, Raycast, Resend, Bun, Stripe's developer products, and GitHub's product pages. Do not copy their branding, layouts, illustrations, wording, or visual identity — match their level of clarity, finish, polish, restraint, confidence, and usability.


The completion bar

You are an autonomous implementation agent. Do not stop after analysis. Do not return recommendations, wireframes, design commentary, a sample section, or a plan.

Create the files → run the project → look at the result → fix the visual and functional defects → iterate until it is polished.

The task is complete only when the site exists in the repository, runs, and is ready to deploy.

A static page with excellent design beats a flashy page with poor usability. A page that still reads as "a README converted into cards" is a failure regardless of how much CSS it has.


Non-negotiable: the repository is the source of truth

Do not rewrite, summarize, restyle, or expand the README. It is one input among many, and usually the least current one.

  • README conflicts with the implementation → trust the implementation.
  • Documentation is outdated → describe what the code does now.
  • A documented feature is missing, partial, stubbed, mocked, disabled, or experimental → qualify it honestly.
  • A real capability exists in code but not in the docs → elevate it onto the page. This is the single most common thing a project under-sells.

Never mistake interfaces for implementations, mocks for working integrations, tests for production features, TODOs for roadmap commitments, architecture placeholders for completed systems, example code for supported behavior, or planned features for available ones.


Phases

Work them in order. Each reference is a working checklist — open it when you reach that phase, not before.

PhaseWhat happensReference
1. InvestigateReverse-engineer the real product: identity, audience, capability inventory, hidden value, maturityreferences/investigation.md
2. Assess the stackFind any existing site, design system, or brand. Choose the lightest stack. Decide where the site livesreferences/stack.md
3. PositionCategory, audience, problem, differentiator, value proposition, capability themes, two CTAsbelow
4. DesignDefine the design system before implementing. Derive it from the product, not from a templatereferences/design.md
5. BuildSection structure, scroll narrative, visual rhythmreferences/structure.md
6. VerifyRun it, inspect it at three widths, fix, repeat. Then the full reviewreferences/review.md
7. READMERewrite the root README as a short entry point that links to the sitebelow

[!IMPORTANT] Phase 2 comes before any design work for a reason. Use the repository's existing design system as the visual identity, so the page looks like the product. Inventing a palette when the product already has one is the difference between a page that looks like the software and a generic template with the right words on it.


Phase 3: positioning

From repository evidence, settle one of each before writing copy, so the page has one clear message:

  • one product category
  • one primary audience
  • one problem statement
  • one strong differentiator
  • one concise value proposition
  • three to six core capability themes
  • one primary call to action, one secondary

The first screen must answer: What is this? Who is it for? Why should I care? Why is it different? Can I trust it? What do I click next?

Avoid powerful, revolutionary, cutting-edge, seamless, robust, next-generation, future-proof, enterprise-grade, all-in-one, best-in-class, game-changing. Each one is a fact you haven't written yet.

Weak: A powerful platform for modern teams. Strong: Run the entire workflow locally, automate it through a documented API, and add new providers without changing the core application.

Translate architecture into consequence. "Uses an adapter architecture" is a note to yourself; "new providers can be added behind a stable interface without rewriting core workflows" is the value.


Evidence classification

Classify every important claim internally before it reaches the page. The tier chooses the verb.

TierMeaningLanguage
ConfirmedImplemented, backed by code, tests, config, or current docssupports · includes · provides · exposes · runs · integrates
Partial / experimentalPresent but incomplete, unstable, gated, or lightly testedincludes experimental support for · provides partial support for · is under active development · currently supports a limited version of
PlannedOnly in roadmap material, issues, specs, or placeholderslabel it planned; never present it as available
UnverifiedSuggested but not adequately supportedexclude it, or describe it cautiously

Never invent: customer counts, usage numbers, star counts, testimonials, benchmarks, performance claims, adoption claims, uptime, reliability or security guarantees, production deployments, enterprise readiness, compatibility, roadmap dates, release timelines, customer logos, awards, or press mentions.

Pick the maturity word the evidence earns — production-ready, production-oriented, actively developed, functional but evolving, early-stage, experimental, prototype, proof of concept — judged on test depth, release history, install reliability, upgrade paths, migrations, error handling, security controls, observability, backups, deployment automation, live TODOs, unfinished UI, and placeholder integrations. Polish is not maturity.


Voice

Confident, human, technically literate. Clear, specific, concise, credible, honest.

Avoid buzzwords, clichés, hype, fake urgency, exclamation marks, generic startup language, vague promises, fake social proof, repetitive sentence patterns, AI-sounding filler, inflated adjectives, and unsupported claims.

Prefer concrete verbs, named capabilities, explicit workflows, real constraints, actual supported platforms, actual deployment methods, actual technical differentiators, and measurable facts where they exist. Every section earns its place.


Links and assets

Verify every link before it ships. Use real repository-relative or public URLs — documentation, installation, architecture, releases, issues, discussions, contribution guide, license, demo, package registry, container registry, hosted site.

No dead placeholder links in the rendered page. When a resource doesn't exist, omit it or leave a code comment for maintainers.

Prefer real project evidence for imagery, in this order: actual product screenshots → repository diagrams → diagrams generated from the real architecture → real code examples → domain-specific illustration → restrained abstract graphics. Never stock photography, never fabricated screenshots, never imagery unrelated to the software. If the app runs locally, run it and capture current screenshots — and check they contain no secrets, tokens, or personal data.


Phase 7: the README

Write it after the site, and keep it short. It is a technical entry point, not a second copy of the website.

# Project Name

One-paragraph product explanation.

[Website] · [Documentation] · [Quick Start] · [Contributing]

## What it does
Short capability summary.

## Quick start
Verified commands.

## Documentation
Links to deeper guides.

## Project status
Honest maturity statement.

## Contributing
Link to the contribution guide.

## License
The actual license.

Link prominently to the landing page. Never overwrite an existing README without saying so; if the repository is a shared clone or worktree, write outside it and explain why.


Deliverables

Where applicable, the finished work includes:

the landing-page website · responsive navigation · hero · product positioning · capability sections · screenshots or visual product proof · workflow explanation · architecture explanation · use cases · differentiation · quick start · deployment information · honest status and limitations · contribution section · final CTA · footer · a design system · reusable components · responsive styles · accessibility support · working build scripts · a deployment workflow · an updated README · docs for running and maintaining the site.


Completion report

Close with a concise report — not a file-by-file narration:

  • what was created, and where the site lives
  • which framework or stack, and why
  • how to run it, build it, and deploy it
  • which repository evidence shaped the positioning
  • which claims were qualified as partial, experimental, or planned
  • any assets or information still needed from maintainers
  • any limitation you could not resolve

Final tests

  1. Does this look and behave like a real world-class product website, or like a README converted into cards? If the latter, redesign it.
  2. Does it use actual product evidence, or generic marketing patterns? If generic, revise.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.