skills/website-cloner/website-implementation-plan/SKILL.md
Turn approved prd.md into a phased implementation plan with landing page first, asset collection vs creation, individual tasks. Writes tasks.md after user approval. Use when asked to plan website implementation, create an implementation plan, or break down a PRD into tasks. Don't use for building or coding — that's a separate phase.
npx skillsauth add luongnv89/skills website-implementation-planInstall 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.
Turns an approved improvement proposal (prd.md) into a phased implementation plan. Landing page first, then deeper pages. Asset collection vs. creation tracked. Writes tasks.md after approval.
Trigger when the user asks to:
Do not use for building or coding — that is Phase 5 (website-builder).
1. Read the approved prd.md
2. Identify phases: landing page first, then deeper content
3. For each task: define scope, outputs, acceptance criteria
4. Track assets: collect from original vs. create new
5. Assemble into tasks.md
6. Present for review
7. Incorporate edits (loop until approved)
8. Persist tasks.md
# Implementation Plan: <site name>
**Source PRD:** prd.md
**Date:** <date>
---
## Overview
Brief summary of the implementation approach and phase ordering rationale.
## Phase 1: Landing Page
The landing/home page is built first so it can be shown to potential users early.
### Task 1.1: Project Setup
**Scope:** Initialize Vite + React + shadcn/ui + Tailwind CSS project. Configure build, routing, and deployment to GitHub Pages.
**Outputs:** Working project scaffold with build pipeline.
**Acceptance Criteria:**
- `npm run dev` starts a local dev server
- `npm run build` produces static assets in dist/
- GitHub Pages deployment configured
**Assets Needed:**
- [Collect] Logo, brand colors, brand name from original site
- [Create] Project repository on GitHub
### Task 1.2: Landing Page Layout
**Scope:** Build the hero section, nav, CTA, and footer based on the improvement proposal. Implement the improved layout, not a clone.
**Outputs:** Landing page component with improved structure.
**Acceptance Criteria:**
- Hero section with clear headline, subtext, primary CTA above the fold
- Responsive layout (mobile + desktop)
- Navigation matches the improved structure
**Assets Needed:**
- [Collect] Hero imagery, copy text, brand colors
- [Create] New CTA copy (if improvement proposes different messaging)
---
## Phase 2: Core Pages
Deeper pages beyond the landing page.
### Task 2.1: <Page Name>
**Scope:** ...
**Outputs:** ...
**Acceptance Criteria:** ...
**Assets Needed:**
- [Collect] ...
- [Create] ...
---
Repeat for each task across phases.
## Phase 3: Optimization and Polish
Performance, SEO, and security improvements from the PRD.
### Task 3.1: Performance Optimization
**Scope:** Image optimization, lazy loading, code splitting, font optimization.
**Acceptance Criteria:**
- LCP ≤ target (from prd.md metrics table)
- CLS ≤ target
- Page weight ≤ target
### Task 3.2: SEO Implementation
**Scope:** Meta tags, structured data, heading structure, alt text, canonical URLs.
**Acceptance Criteria:**
- SEO score ≥ target
- All pages have title, meta description, structured data
### Task 3.3: Security Hardening
**Scope:** HTTPS enforcement, security headers, mixed content fixes.
**Acceptance Criteria:**
- All resources loaded over HTTPS
- Key security headers present
## Asset Summary
| Asset | Source | Action |
|-------|--------|--------|
| Logo | Original site | Collect |
| Brand colors | Original site | Collect |
| Hero image | Original site | Collect |
| CTA copy | Improvement proposal | Create |
| New icons | Generated | Create |
## Deployment
1. Push to GitHub repository
2. Configure GitHub Pages (settings → Pages → source: main /docs or /)
3. Verify deployment at `https://<user>.github.io/<repo>/`
---
*This plan is derived from the approved improvement proposal. Actual task scope may need adjustment during implementation.*
Read file <path-to-prd.md>
If missing, ask for the path. The orchestrator should have produced this in Phase 3.
Structure phases so something usable ships early:
| Phase | Focus | Rationale | |-------|-------|-----------| | Phase 1 | Landing/home page | Usable immediately, can be shown to users | | Phase 2 | Core pages | About, features, contact, etc. | | Phase 3 | Optimization | Performance, SEO, security polish | | Phase 4 (optional) | Extra features | Nice-to-have improvements |
Phase 1 must produce an independently usable landing page.
For each task:
[Collect] from [Create]Tasks should be small enough for a single implementation cycle.
For every asset referenced in the plan:
Assemble using the structure above.
"Here is the implementation plan. Please:
If edits requested: update, re-present, repeat until approved.
Do not persist until explicit approval.
printf '%s\n' "$TASKS_CONTENT" > "$OUTPUT_PATH"
Default: $PROJECT_DIR/tasks.md or ~/workspace/clones/YYYY_MM_DD_slug/tasks.md.
If $ARGUMENTS includes --output <path>, use that.
Confirm:
tasks.md saved to: <absolute-path>
STATUS: approved
When invoked by the website-cloner umbrella (Phase 4 gate), the orchestrator
gates Phase 5 on this skill's outcome. The contract:
| Outcome | Signal |
|----------|---------------------------------------------------------------------|
| approved | tasks.md exists at the resolved output path AND final line of stdout reads STATUS: approved |
| pending | no tasks.md written; final line reads STATUS: pending (user still iterating) |
| aborted | no tasks.md written; final line reads STATUS: aborted (user declined) |
The orchestrator MUST NOT advance to Phase 5 unless the outcome is approved.
A standalone invocation may ignore the status line, but the file-existence rule
still holds: no approval, no tasks.md.
| Failure | Behavior | |---|---| | No prd.md provided | Ask for the PRD file path | | Invalid PRD format | Report error and ask for valid file | | User never approves | Keep looping; do not auto-save |
documentation
Manage software releases end-to-end: bump version, generate changelog, tag, push, GitHub release, publish to PyPI/npm. Use when user asks to ship, cut a release, tag a version, or list changes since last tag. Skip routine commits and marketplace publishing.
development
Review UI for usability issues using Steve Krug's principles and produce a scannable report. Use when asked for a usability audit, UX review, or UI feedback on screenshots, URLs, or code. Don't use for visual/brand design critique, accessibility (WCAG) audits, or backend/API review.
development
Validate app/startup ideas with market, feasibility, commercial, and open-source competitor analysis. Use when asked to evaluate, validate, or score a product idea. Don't use for PRDs, go-to-market plans, or investor decks.
testing
Install local-first security hardening: pre-commit secret detection, offline dependency scans, static analysis, reports, and gated free CI. Use when hardening repos or adding security hooks. Don't use for incident response or cloud security reviews.