skills/readme-to-landing-page/SKILL.md
Transform a project README.md into a conversion-optimized landing page using PAS, AIDA, or StoryBrand — hero, value prop, how it works, quick start, CTA in pure markdown. Skip for full HTML sites or general blog copy.
npx skillsauth add luongnv89/skills readme-to-landing-pageInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Transform a project's README.md into a concise, visual, developer-friendly landing page.
Developers skim — they don't read walls of text. The output README must:
If something can be a diagram, make it a diagram. If something can be a table, make it a table. If something can be a one-liner, don't write three sentences.
AI-generated READMEs have predictable tells. Slop signals "nobody reviewed this" and kills credibility. See references/anti-slop-rules.md for the full banned-phrase list and structural patterns to avoid.
Quick test: read each sentence and ask "does this give the reader information they didn't already have?" If not, cut it.
branch="$(git rev-parse --abbrev-ref HEAD)"
git fetch origin
git pull --rebase origin "$branch"
If the working tree is dirty: git stash push -u -m "pre-sync", sync, then git stash pop.
If origin is missing or conflicts occur, stop and ask the user.
If no README.md exists, or the existing one is already marketing-style, see Edge Cases below before starting Step 1 — a different mode applies.
Read project files to build a picture:
| File | Purpose |
|---|---|
| README.md | Source material (required) |
| package.json / pyproject.toml / Cargo.toml / go.mod | Name, description, version |
| CHANGELOG.md | Momentum signals |
| LICENSE | Risk reversal ("MIT licensed") |
| docs/ | Existing detailed docs to preserve |
Also check: GitHub stars, forks, download stats, notable users. If purpose or audience is unclear, ask.
Determine these three things — write them down before proceeding:
| Project Type | Framework | Approach | |---|---|---| | CLI tools, libraries, infra | PAS | Problem -> Agitate -> Solution | | Apps, platforms, end-user products | AIDA | Attention -> Interest -> Desire -> Action | | Strong narrative / user request | StoryBrand | Hero -> Guide -> Plan -> CTA |
Tell the user which you picked. Let them override.
cp README.md README.backup.md
Follow the section flow in references/section-templates.md (hero, how it works, features, quick start, examples, comparison, social proof, FAQ, CTA, collapsed technical details). Each template includes good vs. bad patterns and exact length budgets.
Key rules across all sections:
<details> blocksBefore presenting to the user, verify:
| Check | Rule |
|---|---|
| Code blocks | Each independently-runnable command is in its own block |
| Mermaid diagrams | >=1 diagram showing how the project works |
| Hero length | <=5 lines (excluding badges) |
| Feature descriptions | <=15 words each |
| Quick Start | <=5 commands |
| Paragraphs | No paragraph exceeds 3 sentences |
| Total prose | Diagrams + tables + code >= 50% of content |
| Original content | All preserved in <details> blocks |
| Fabricated data | None — real numbers or placeholders only |
| No emoji | Zero emoji anywhere in the output |
| No slop phrases | None of the banned phrases from references/anti-slop-rules.md |
| No filler | Every sentence adds information the reader didn't have |
| No rhetorical questions | No "Tired of X?" or "Want to Y?" patterns |
If any check fails, fix it before presenting.
Briefly explain:
README.backup.md + collapsed sections)Ask for feedback. Do NOT commit unless asked.
For a CLI tool called "fastbuild", the rewritten README.md opens with:


# Incremental builds in under a second
fastbuild watches your source tree and recompiles only changed modules —
no config files, no daemon, no warm-up time.
[**Get Started ->**](#getting-started)
## How It Works
` ` `mermaid
graph LR
A[Source files] --> B[fastbuild watcher]
B --> C{Changed?}
C -->|Yes| D[Compile changed modules]
C -->|No| E[Skip]
D --> F[Output bundle]
` ` `
| Feature | What you get |
|---|---|
| Zero config | Works out of the box — no setup files |
| Incremental | Recompiles only changed modules |
| Fast | < 1s on codebases with 10k+ files |
Original content is preserved in README.backup.md and in <details> blocks at the end of the new README.
<details> blocks; landing-page sections stay brief.README.backup.md; summarize technical sections as collapsed <details> blocks.README.backup.md before any rewriteAfter each major step, output a status report. See references/step-reports.md for the format and phase-specific checks.
[placeholder] markers.tools
Run Herdr loops for one open GitHub issue (resolve→review→fix) or an existing PR (review→lazy fixer) until CLEAN. Don't use for plain resolution without review, review-only/no-fix requests, backlog automation, or merging.
tools
Manage AI agent fleets in Herdr: split root + sub-agents into one tab as a tiled grid, message/wait/read via herdr CLI, steer any pane. Use for Herdr multi-agent fleets. Don't use for tmux, screen, or non-Herdr terminals.
development
Generate or update docs to match the code, citing each claim to path:line and asking on ambiguity; runbook docs also get a check-only validation script. Don't use for API-reference autogen (JSDoc/Sphinx), landing pages, or CLAUDE.md/AGENTS.md.
testing
Generate a diagram and route to the right engine — draw.io XML (precise, editable, C4, swimlanes) or Excalidraw JSON (hand-drawn, sketch, wireframes). One entry for flowcharts, architecture, ER, sequence, mind maps. Don't use for Mermaid or slides.