尚未偵測到更新。GitHub 發布版本後會顯示在這裡。
個人 GitHub 帳號下公開儲存庫的擁有者可以登入並驗證所有權,自行管理介紹,無需管理員核准。
登入並產生驗證檔案,在 30 分鐘內將其提交至此儲存庫的預設分支。其他分支或分叉儲存庫中的檔案無效。然後在下方檢查檔案,即可無需管理員核准直接管理專案。請保留檔案以維持管理權限。
如需登入,將前往登入頁面。登入後請回到此處,再次點擊所選的驗證按鈕。
在專案資料夾中執行 /nomorevibe verify,即可驗證所有權並自行管理介紹。
一分鐘完成驗證製作者的介紹僅有上面的一句話。 更多詳情請閱讀儲存庫 README。 儲存庫 README ↗
正在載入完整 README。
檢視完整 README ↗A lean starting point for client migrations, rebuilds, and new startup sites. The template reuses launch discipline—not a runtime page builder.
docs/launch/checklist.md. Its source and current state live in docs/launch/checklist.json.docs/features.md during project kickoff. Default features are normal code: delete them when unused instead of adding flags.docs/launch/migration.md and docs/launch/url-map.csv.pnpm launch:checklist --set <ID> <todo|done|not_applicable>.pnpm launch:checklist --add-project <ID> <P0|P1|P2> <title> --detail <Markdown>.PLAN.md for the architecture audit and page-by-page roadmap.Requirements: Node.js 20.9+ and pnpm 10.
pnpm install
cp .env.example .env.local
pnpm dev
Quality checks:
pnpm launch:checklist
pnpm check
pnpm build
pnpm launch:audit --url http://localhost:3000 --mode template
Launch commands have separate jobs:
pnpm launch:checklist displays manual progress and identifies automated items. Its --set, --add-project, --write, and --check options manage or validate the canonical checklist.pnpm launch:verify executes every named check whose checklist status is auto. It needs no running website and exits non-zero when a check fails.pnpm launch:audit runs those automated checks plus live crawling, route, metadata, indexability, sitemap, and production-safety checks against a URL.pnpm indexnow:key creates or rotates a site's IndexNow ownership key and root-level public verification file.pnpm check includes pnpm launch:checklist --check so invalid checklist data or stale generated Markdown fails normal CI. It does not run pnpm launch:verify, because the base template intentionally contains marked placeholders until a client replaces them.
Before production, configure the real identity and run:
pnpm launch:verify
pnpm launch:audit --url https://example.com --mode production
pnpm launch:verify runs every checklist item whose status is auto; it exits non-zero while any named check fails. The production audit runs those functions alongside live route, metadata, indexability, and sitemap checks. It also fails while any manual P0 item remains todo.
/contact action, a stack trust bar, shared positioning sentence, reordered existing sections, FAQ, and close. Customer-proof sections stay omitted until real inputs exist; see docs/recipes/homepage.md.base-nova style; only the Button, Card, and Badge primitives used by current pages are checked in.src/instrumentation-client.ts; add its token to activate it or delete the file and dependency./use-cases and /industries hubs, plus one detail page per JSON file in src/content/pages/{uses,industries}. Feature ids resolve through the shared visual resolver; see docs/recipes/use-cases.md./dev, the full /dev/visuals inventory, and three template launch frames at /dev/launch-assets. pnpm launch:assets builds the production app and exports exact reduced-motion PNGs to artifacts/launch-assets; see docs/recipes/launch-assets.md.llms.txt: a complete website-level example selling Launch Template, conditional homepage Link discovery header, focused source validator, production header/response/public-target audit, and configuration/removal recipe are included. Deleting public/llms.txt removes the discovery header rule on the next build or server start. pnpm launch:verify deliberately fails until the example is replaced for the client.Accept: text/markdown, and /path.md returns the exact same representation for clients that cannot set headers. HTML remains the default on canonical URLs, quality values and HTTP 406 are handled, and optional files under can override individual paths. See .src/lib/seo.ts.src/components/ui; add a component only when a current page uses it.site.json, a marketing catch-all renderer, or a block registry.$site-clone inventories an authorized source site, maps exact URLs and wildcard page families, migrates original assets, reproduces visible widgets, journals functional gaps, and drives a screenshot/diff loop through homepage approval before other routes.$micro-ui builds route-native interface visuals, route-local motion, local use-case images, generated social images, and browser-exported launch assets. Looping capability visuals use its shared framer-motion kit; page-level motion stays CSS-first. Server-rendered content stays visible without JavaScript and every animation has a composed reduced-motion state. Reusable visuals resolve through one bounded typed map shared by the detail pages, the gated inventory, and launch assets instead of a site-wide block registry.The .claude/skills/* entries are symlinks to .agents/skills/*. Keep the names aligned and validate both the canonical skill and its symlink after changing one.
The template code is available under the MIT License.
Global identity and navigation live in src/config/site.ts; src/config/env.ts parses the deployment-specific site URL and Wisp publication ID. Page content does not belong in global config.
Set the server-only WISP_BLOG_ID to the publication used by /blog. Featured content and optional filters are declared in src/app/blog/blog.config.ts. See docs/recipes/blog.md for configuration and complete removal steps.
Set all MAIL_* values and CONTACT_TO_EMAIL to activate /contact; none may use a NEXT_PUBLIC_ prefix. The default stores only first-touch campaign fields and five recent same-site paths, and sends no form values to PostHog. See docs/recipes/contact.md for privacy decisions, verification, and removal.
Set both NEXT_PUBLIC_TURNSTILE_SITE_KEY and server-only TURNSTILE_SECRET_KEY to add Cloudflare Turnstile to the contact form. With either value absent, the widget script is not loaded and server verification is skipped. See docs/recipes/turnstile.md for the reusable file map and setup flow.
Set server-only VISUAL_REVIEW_ENABLED=true to expose /dev, /dev/visuals, and /dev/launch-assets. .env.example enables it for copied local configuration; do not set it on a public deployment. The exporter supplies it only to its temporary local server.
Markdown responses require no environment values and normally require no duplicate content file. Add src/content/markdown/index.md for a homepage override or mirror another public pathname under src/content/markdown only when its agent-facing representation should differ. See docs/recipes/markdown-negotiation.md.
robots.txt always allows public routes and publishes the sitemap URL. It
blocks only /api/ and /dev/. Protect private previews with deployment
access controls instead of a site-wide robots block.
src/content/markdowndocs/recipes/markdown-negotiation.md$site-clone and $micro-ui live in .agents/skills. Claude discovers the same files through .claude/skills symlinks, so edit only the canonical .agents copies.