跳到正文
cccskills
免费在 GitHub 公开

build-daily-inspiration-sites

Turn five collected UI references into five original landing pages with distinct brands, copy, and imagery. Each page gets its own Codex task and workspace, with build checks, a source commit, and private deployment through Sites.

中文简介正在准备中。当前显示英语简介。

原文查看原文说明

Turn a completed daily UI inspiration capture into exactly five original landing-page builds, one per separate Codex task, using Sites. Use when the user asks to turn the daily inspiration references, a five-item UI prompt pack, or a dated `*-ui-inspiration-capture` article into distinct HTML landing pages while changing the source brands, names, copy, people, numbers, pricing, claims, and imagery.

实测评分 未实测

查看安装方法

适用场景

  • Build five landing pages from a UI inspiration pack
  • Explore distinct brands using collected visual references
  • Manage each page in a separate Codex task
  • Validate and privately deploy five Sites projects

包含的文件(4)

  • SKILL.md7.4 KB
  • agents/openai.yaml292 B
  • references/thread-brief-contract.md5.2 KB
  • scripts/validate_capture.mjs2.6 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

Build Daily Inspiration Sites

Convert an existing five-item inspiration pack into five independent, original Sites projects. Orchestrate from the current task; let each created task own one build, commit, and private deployment.

Preserve the boundary

  • Use the capture as evidence and art direction, never as a template to copy.
  • Do not recollect inspiration unless the user explicitly asks.
  • Create exactly one site for each manifest item and exactly one Codex task for each site.
  • Do not combine multiple sites in one build task.
  • Do not reuse the reference brand, logo, wordmark, copy, people, companies, testimonials, figures, dates, plans, prices, URLs, legal text, interface screenshots, photography, or proprietary marks.
  • Preserve only high-level visual grammar visible in the full-page image: hierarchy, pacing, section logic, contrast, spacing, and image treatment.
  • Attach exactly one reference image to each build task: that item's full-page image. Do not attach a representative still, video, motion frames, or section crops.
  • Use the Codex in-app browser only. Never use Chrome.

1. Locate and validate the capture

Use the article path supplied by the user. Otherwise choose the latest dated articles/YYYY-MM-DD-ui-inspiration-capture/ folder in the current Content workspace.

Require:

  • manifest.json with itemCount: 5 and exactly five items
  • content.md
  • one non-empty local fullPageImage per item
  • pageUrl for every live-site item
  • one detailed builder prompt per item

Run:

node <skill-dir>/scripts/validate_capture.mjs <article-dir>/manifest.json

Stop if validation fails. Report a missing full-page image rather than substituting another still, video, frame, crop, or live-page capture.

2. Create the originality matrix

Before creating tasks, define five distinct replacement identities. For each item, choose:

  • a new brand name and slug
  • a new category-specific positioning statement
  • a fresh audience and conversion goal
  • a new content vocabulary
  • a new set of proof points, metrics, plans, and prices
  • a new imagery direction appropriate to the concept

Keep the new names materially different from the reference titles. Search the five proposed names against one another for accidental similarity. Avoid five generic AI names or five cosmetic color swaps.

The transformation must be coherent. A new statistic must agree everywhere it appears; pricing, testimonials, locations, service names, product UI, and footer details must all belong to the replacement brand.

3. Prepare one build brief per item

Read references/thread-brief-contract.md before creating any task. Fill every placeholder with concrete values; never dispatch a brief containing NEW_BRAND, TODO, or other unresolved markers.

Each brief must include:

  • the exact manifest item title and creator
  • absolute paths to manifest.json and content.md
  • exactly one inline local-image attachment using the item's absolute fullPageImage path in non-fenced Markdown image syntax: ![Full-page reference — <REFERENCE_TITLE>](<ABSOLUTE_FULL_PAGE_IMAGE_PATH>)
  • the new brand name, category, audience, offer, and conversion goal
  • an explicit originality boundary and exhaustive replacement list
  • the requirement to use sites:sites-building and sites:sites-hosting
  • one semantic, one-route HTML landing page; the Sites starter may render it through app/page.tsx
  • full responsive, accessibility, reduced-motion, and interaction requirements
  • build validation, narrow commit, private deployment, and final URL requirements

The Markdown image must render as an attachment in the created task. Do not provide the full-page path as plain text, and do not attach any other image or video. If the task-creation surface cannot render the local Markdown image, stop and report that blocker instead of falling back to a path-only handoff.

Tell each task to inspect the attached full-page image before coding. The detailed prompt in the manifest or article is the section-by-section source of truth. A static full-page image does not establish source motion, interaction behavior, or implementation technique, so do not infer or claim those details.

4. Create five separate Codex tasks

Use Codex thread tools only when the user explicitly requested separate tasks or threads.

  1. Search recent tasks for the article date plus reference title or proposed brand name. Reuse an existing matching task instead of creating a duplicate.
  2. Prefer a projectless target with a unique directory name for a new independent Sites project.
  3. If the user requires the source inside an existing repository, list projects first and create a separate worktree for each build.
  4. Create all five tasks with their complete briefs.
  5. Give each task a clear title such as Build <Brand> landing page.
  6. Confirm all five tasks are active. Use the thread wait/status tool; if waiting is unavailable, use the task list as the fallback proof.

Never edit the five sites concurrently in the Content checkout. The capture files are read-only evidence; each build writes only to its own task workspace.

If task creation is unavailable, stop and report that blocker. Do not silently collapse the work into the current task.

5. Require a complete Sites handoff

Every created task must:

  • initialize Sites immediately
  • build a complete single-route landing page rather than a hero mockup
  • use semantic structure and original, appropriately licensed media
  • avoid model-authored SVG illustrations
  • run the required production build and fix failures
  • commit only the validated site source
  • publish privately through Sites after a successful build
  • follow Sites approval rules if only shared or public access is available
  • return the deployed URL, source location, commit, and concise change summary

Because these are background tasks, skip interactive browser preview unless it is needed to resolve a real blocker. Publishing and the normal Sites handoff still apply.

6. Report the launch

Return a compact mapping of reference to replacement brand and current task status. Emit one created-task directive per new task when the app requires it.

State clearly that the five tasks will each produce:

  • one original landing page
  • one isolated source workspace
  • one validated commit
  • one private Sites deployment

Do not claim a site is complete until its task reports successful build and deployment.

Completion checks

  • Exactly five manifest items were validated.
  • Exactly one non-empty full-page image was validated for each item.
  • Exactly five distinct replacement identities were defined.
  • Exactly five unique tasks were created or deliberately reused.
  • Every task contains exactly one rendered full-page image attachment and no path-only image handoff.
  • Every brief replaces all source identity, language, numbers, and imagery.
  • Every task requires Sites build, validation, commit, and private deployment.
  • The user can open each created task independently.

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

Build woven fabric in Three.js with visible threads, realistic sheen, seams, and thick hems. The skill covers folds that preserve texture proportions, light passing through thin cloth, and variations in dye and wear.

中文简介正在准备中。当前显示英语简介。

  • Creating backlit curtains
  • Differentiating linen, silk, and wool
  • Adding hems and stitches for close-up views
MengTo/Skills69792026年10月18日 更新

Build falling leaves in Three.js and similar 3D scenes, with visible front, edge, and back views as they tumble and drift. Coordinate wind, scene depth, camera-aware recycling, and performance checks for a convincing autumn effect.

中文简介正在准备中。当前显示英语简介。

  • Adding falling autumn leaves to a 3D scene
  • Coordinating leaf tumble, sideways drift, and wind
  • Managing leaves around buildings and camera movement
MengTo/Skills69792026年10月18日 更新

Add coordinated seasonal transitions to an existing 3D scene. Blend foliage, lighting, sky, ground, snow, and particles from one shared state while preserving camera controls and interactions, so the same place remains recognizable throughout spring, summer, fall, and winter.

中文简介正在准备中。当前显示英语简介。

  • Add a four-season selector to an outdoor 3D scene
  • Show seasonal changes in an architectural walkthrough
  • Coordinate foliage, sunlight, sky, and ground appearance
MengTo/Skills69792026年10月18日 更新

Create or integrate detailed 3D models with smooth outlines and believable edges for close-up views. Prepare lower-detail versions for distant viewing, prioritize loading, and measure rendering costs in the actual scene.

中文简介正在准备中。当前显示英语简介。

  • Creating a detailed 3D model for close-up views
  • Replacing visibly faceted curves with smoother geometry
  • Preparing detailed architectural assets for rendering
MengTo/Skills69792026年10月18日 更新

Prepare detailed 3D materials that remain clear in close-up views. The skill guides texture resolution, physically based material maps, surface mapping, filtering, and progressive loading while checking download size and GPU memory use.

中文简介正在准备中。当前显示英语简介。

  • Prepare wood, stone, and fabric materials for close-up views
  • Fix texture seams, grain direction, and normal map orientation
  • Preserve floor and foliage detail at oblique or distant views
MengTo/Skills69792026年10月18日 更新

Create realistic metal surfaces in Three.js by combining structured reflections, studio lighting, scratches, fingerprints, and beveled edges. Includes polished, brushed, and aged finishes for objects such as tools, rings, and metal product parts.

中文简介正在准备中。当前显示英语简介。

  • Fixing metal that looks like gray plastic
  • Creating polished or brushed metal finishes
  • Adding patina and worn edges
MengTo/Skills69792026年10月18日 更新

查看 MengTo 的全部技能

报告此技能的问题