@getvitops/generator 8.2.0-alpha.0 → 8.3.0-alpha.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/CHANGELOG.md +550 -25
- package/assets/css/index.css +7 -0
- package/assets/css/patterns/banner.css +20 -0
- package/assets/css/patterns/callout.css +39 -0
- package/assets/css/patterns/copy.css +5 -0
- package/assets/css/patterns/embed.css +39 -0
- package/assets/css/patterns/forms.css +19 -0
- package/assets/css/patterns/logos.css +45 -0
- package/assets/css/patterns/marquee.css +16 -0
- package/assets/css/patterns/section.css +53 -0
- package/assets/css/patterns/sitenav.css +262 -0
- package/assets/css/patterns/splitter.css +11 -0
- package/assets/css/patterns/stats.css +61 -0
- package/assets/css/patterns/steps.css +98 -0
- package/assets/css/patterns/table.css +10 -0
- package/assets/css/patterns/tone.css +38 -0
- package/assets/js/consent.js +1 -1
- package/assets/js/elements.js +5 -5
- package/assets/js/polyfills/adoptedStyleSheets.js +1 -1
- package/assets/js/polyfills/css-anchor-positioning-fn.js +2 -2
- package/assets/js/polyfills/popover.js +2 -2
- package/config.schema.json +257 -120
- package/dist/config-BfcWKH-5.mjs +1743 -0
- package/dist/config-BfcWKH-5.mjs.map +1 -0
- package/dist/{config-DvbYx3Gi.mjs → config-Bhg3Xv3D.mjs} +10 -207
- package/dist/config-Bhg3Xv3D.mjs.map +1 -0
- package/dist/config-D5xuz8--.mjs +1807 -0
- package/dist/config-D5xuz8--.mjs.map +1 -0
- package/dist/config.d.mts +157 -211
- package/dist/config.d.mts.map +1 -1
- package/dist/config.mjs +1 -1
- package/dist/{icons-sprite-Bf_KmKeX.d.mts → icons-sprite-CQ5Zj4go.d.mts} +4 -6
- package/dist/icons-sprite-CQ5Zj4go.d.mts.map +1 -0
- package/dist/icons-sprite.d.mts +1 -1
- package/dist/icons-sprite.mjs +1 -2
- package/dist/index.d.mts +49 -35
- package/dist/index.d.mts.map +1 -1
- package/dist/index.mjs +607 -75
- package/dist/index.mjs.map +1 -1
- package/dist/{schema-Dmq1J7S3.d.mts → schema-BV2n2phj.d.mts} +2 -3
- package/dist/{schema-Dmq1J7S3.d.mts.map → schema-BV2n2phj.d.mts.map} +1 -1
- package/dist/schema-C-P9dF9F.d.mts +337 -0
- package/dist/schema-C-P9dF9F.d.mts.map +1 -0
- package/dist/{tokens-8SXMDAd1.d.mts → tokens-BzBbTeTi.d.mts} +2 -6
- package/dist/tokens-BzBbTeTi.d.mts.map +1 -0
- package/dist/{tokens-zi594IZ3.mjs → tokens-C1RJwFeG.mjs} +3 -5
- package/dist/{tokens-zi594IZ3.mjs.map → tokens-C1RJwFeG.mjs.map} +1 -1
- package/dist/tokens-CaIGQQLk.mjs +740 -0
- package/dist/tokens-CaIGQQLk.mjs.map +1 -0
- package/dist/tokens-DH5cpvKd.d.mts +242 -0
- package/dist/tokens-DH5cpvKd.d.mts.map +1 -0
- package/dist/tokens.d.mts +1 -1
- package/dist/tokens.mjs +1 -1
- package/package.json +6 -5
- package/schema.json +95 -5
- package/dist/config-DvbYx3Gi.mjs.map +0 -1
- package/dist/icons-sprite-Bf_KmKeX.d.mts.map +0 -1
- package/dist/tokens-8SXMDAd1.d.mts.map +0 -1
package/CHANGELOG.md
CHANGED
|
@@ -1,5 +1,538 @@
|
|
|
1
1
|
# @getvitops/generator
|
|
2
2
|
|
|
3
|
+
## 8.3.0-alpha.0
|
|
4
|
+
|
|
5
|
+
### Minor Changes
|
|
6
|
+
|
|
7
|
+
- 93d2067: New framework patterns and an `Entries` component for the EmDash block library.
|
|
8
|
+
|
|
9
|
+
**CSS, all formats** (css, bricks and inlined into tailwind's `components` layer; no existing class changes): `tone` (a shared tint/solid colour convention driven by `--tone-*` custom properties, plus `.accent` for a run inside a heading), `section` (head and heading-left `--split` layout), `steps` (numbered `<ol>` with a connector line), `stats` (figures and `--tiles`), `callout` (a call-to-action band), `logos` (a still partner grid) and `embed` (a consent-gated iframe frame). All are structural partials, classified `vitops.components`.
|
|
10
|
+
|
|
11
|
+
**`@getvitops/astro`:** new `Entries.astro` (`@getvitops/astro/components/Entries.astro`) emits `<wc-entries>` itself with the semantic heading + `<dl>` fallback.
|
|
12
|
+
|
|
13
|
+
**`@getvitops/generator`:** `TierEntry` gains `emdashExempt` (a reason a shipped web component has no EmDash block). `tiers.test.ts` now fails if a shipped `wc-*` has neither `emdash` nor `emdashExempt`, and the components doc and the docs site list the exemptions.
|
|
14
|
+
|
|
15
|
+
No other consumer-facing change in `@getvitops/core`, `@getvitops/utils`, `@getvitops/cli` or `@getvitops/vite` beyond the version bump.
|
|
16
|
+
|
|
17
|
+
- 161c7f3: A contact form that works end to end: `vitops setup turnstile`, `<wc-turnstile>`, the `vitops.form` EmDash block, and the route the emdash template ships to receive it.
|
|
18
|
+
|
|
19
|
+
**CSS, all formats** (no existing class changes): `forms.css` gains `.form-notice--target:not(:target)` (a message revealed only when the URL fragment names it, which is how a prerendered page shows a no-JS form's result) and `:where(wc-turnstile) { display: block }`.
|
|
20
|
+
|
|
21
|
+
**`@getvitops/core`:** new `<wc-turnstile sitekey="…">` in `elements.js`. It loads Cloudflare Turnstile on the first interaction with its form (no request to `challenges.cloudflare.com` for a visitor who only reads the page), holds an early submit until the token arrives, and leaves its slotted no-JS explanation in place until JS takes over. Turnstile is category `necessary`, so it is not behind the consent banner.
|
|
22
|
+
|
|
23
|
+
**`@getvitops/utils` / `@getvitops/cli`:** new `@getvitops/utils/turnstile` and `vitops setup turnstile [--dry] [--check] [--no-write] [--no-secret]`. It plans from the site's hostnames, creates a managed widget unless one already covers them (an existing widget is adopted, never edited, rotated or deleted), writes the public key to `site.security.turnstile.siteKey`, and pipes the one-time secret to `wrangler secret put VITOPS_TURNSTILE_SECRET` without writing it to any file. Needs `CLOUDFLARE_API_TOKEN` with **Account > Turnstile > Edit** and `CLOUDFLARE_ACCOUNT_ID`.
|
|
24
|
+
|
|
25
|
+
**`@getvitops/astro`:** `virtual:getvitops/head` gains `turnstileSiteKey` (read from `site.security.turnstile.siteKey`) so blocks can render the widget. No other change.
|
|
26
|
+
|
|
27
|
+
**`@getvitops/generator`:** `TIERS` gains a `turnstile` entry (wc-turnstile, claimed by `vitops.form`); `wc-multi-field`'s `emdashExempt` reason is reworded now the form block exists.
|
|
28
|
+
|
|
29
|
+
**`@getvitops/emdash`:** new `vitops.form` block: a field repeater (text, email, phone, long text, choose one, tick box), required flags, help text, success/error messages, optional Turnstile. It renders a semantic `<form method="post">` that works with no JavaScript, with a honeypot named per `createConversionRoute`'s defaults. A field named `phone` submits as `tel`, because the route reserves `phone` and `event`.
|
|
30
|
+
|
|
31
|
+
**`@getvitops/create`:** the emdash template ships `src/pages/api/track.ts` (`prerender = false`, `createConversionRoute()`), a seeded contact form, and `challenges.cloudflare.com` in both templates' Report-Only CSP (`script-src`, `frame-src`). The route refuses a post with nowhere to deliver and fails closed when a site key exists without `VITOPS_TURNSTILE_SECRET`. SETUP.html gains the Turnstile step. Existing sites: copy the route and the CSP change, then run `vitops setup turnstile`.
|
|
32
|
+
|
|
33
|
+
- 175d16e: **Added: `vitops.reviews` and `vitops.nav` EmDash blocks.** Non-breaking.
|
|
34
|
+
|
|
35
|
+
- **`vitops.reviews`** renders Google reviews from the committed `.vitops/reviews.json` cache (`vitops reviews`) as a grid, carousel or marquee of `<ReviewCard>`. Fields: location key, layout, columns, limit, lowest rating, order, featured ids, reviewer photos (off by default), tone/colour. Which reviews show is `selectReviews()`, the one shared decision. Review JSON-LD is not offered (Google treats self-marked-up reviews as ineligible for rich results). A new review reaches the page after `vitops reviews` is re-run, committed and the site rebuilt, not when an editor saves; with no cache file the block renders nothing (a note in `astro dev`).
|
|
36
|
+
- **`vitops.nav`** renders a named EmDash menu inside page content (site map, section links, footer-style columns), never the header. Fields: menu (picked from the site's menus via a new `menus/list` plugin route, with `footer-sitemap`/`footer-legal`/`primary` as the static fallback), accessible name, visible heading and level, layout `list | inline | columns | sitenav`, depth 1-3, tone/colour. It is a `<nav>` with a name that is never `Primary`, `Site map` or `Legal` (the layout's landmarks), `aria-current="page"`, no runtime JS; a missing or empty menu renders nothing.
|
|
37
|
+
- `@getvitops/astro`: **`virtual:getvitops/reviews`**, served by `vitops()`, is `.vitops/reviews.json` as a build-time import (an empty cache when absent or unreadable), because a Worker cannot read the file. `<ReviewCard>` now renders the review's source link (`.review__source`, `rel="nofollow noopener noreferrer"`), which the pattern already styled. **Fix:** `src/json-ld.ts` was missing from the published `files`, so any consumer importing `<ReviewCard>` (which composes `schemas/Review.astro`) failed to build with an unresolved import. `@getvitops/emdash` now needs this version of `@getvitops/astro` for the `vitops.reviews` block.
|
|
38
|
+
- `@getvitops/utils`: new Worker-safe subpath **`@getvitops/utils/reviews/select`** (`selectReviews` and the review types, no `node:fs`). `selectReviews` now takes only the display fields (`featured`, `minRating`, `limit`, `sort`), so a caller with no configured `locationId` can use it; existing calls are unchanged.
|
|
39
|
+
- `@getvitops/core`: new `.navlist` pattern in `patterns/sitenav.css` (`.navlist`, `__item`, `__link`, `__label`, `__sub`, `--inline`, `--columns`). **All formats** (css, bricks, tailwind). No existing class changes.
|
|
40
|
+
- `@getvitops/generator`: `TIERS.review` lists `vitops.reviews` and `TIERS.sitenav` lists `vitops.nav` (and `.navlist`), so `vitops docs components` and the docsite show them.
|
|
41
|
+
|
|
42
|
+
- 83d6f3c: Theming escape hatches, so a real palette builds without raw ramp steps. All additive; existing configs build unchanged.
|
|
43
|
+
|
|
44
|
+
- **Dark-only sites.** `designSystem.defaultColorScheme: "dark"` (accepted by the schema before, but ignored) now makes the dark tokens the `:root` appearance with `color-scheme: dark`, emits no light block or `data-theme` flip, and skips light-mode contrast checks. Reaches css, bricks and tailwind formats. Do not ship `<wc-color-scheme-toggle>` on such a site.
|
|
45
|
+
- **Role solid step and foreground.** `colors.roles.<role>` may be `{ hue, kind?, solid?, foreground? }`. `solid: 400` starts the solid family at step 400 (bold and x-bold walk one and two rungs darker); `foreground: "dark" | "light"` pins `text-on-<role>`. The APCA check (Lc 60) still runs on the pair emitted. The string shorthand is unchanged. All formats. The live editor's `roleTokens` still assume the default 600 family when re-pointing a role.
|
|
46
|
+
- **Split shadows.** `shadows.<name>` may be `{ box?, drop? }`: `box` (spread, inset, layers) is `--shadow-<name>`; `drop` is `--shadow-<name>-drop` and feeds `.drop-shadow-<name>` and the `shadow:` state shortcut. The string form is unchanged. css/bricks emit `.drop-shadow-*`; tailwind emits the `-drop` token only.
|
|
47
|
+
- **Layout tokens.** New optional `layout` block (`widths.{measure,breakout,spotlight,nav}`, `regionSpace.{min,preferred,max}`) emitted as `--width-*` / `--region-space*` on `:root` in css, bricks and tailwind formats, and exposed in `design-manifest.json` (`layout`, `reverseIndex`). Keys you omit keep the framework defaults.
|
|
48
|
+
- **`acknowledgeOffLadder: true`** on a palette hue silences that hue's off-ladder anchor warnings.
|
|
49
|
+
- **Typography:** `transform` and `decoration` are now accepted as aliases of `text-transform` and `text-decoration` (canonical wins when both are present) instead of warning and being dropped.
|
|
50
|
+
- **Docs strings** (`authoring.md`, JSON Schema hovers) state the Lc 60 muted-text requirement against the page plane and give a type-scale recipe for marketing headings.
|
|
51
|
+
- `@getvitops/utils`: `pickOn` takes an optional `prefer: 'light' | 'dark'`.
|
|
52
|
+
|
|
53
|
+
- 892b204: Generated legal documents gain a **Quebec (Law 25) jurisdiction**: `site.legal.jurisdiction: "qc"` renders the privacy policy, terms of use and cookie notice from new, Quebec-specific templates (not the `ca` prose). The privacy policy adds the person in charge of the protection of personal information, technologies that identify, locate or profile, transfers _outside Quebec_ (so a provider elsewhere in Canada counts), retention and anonymization, and the rights of access, rectification, portability, withdrawal, de-indexing and review of automated decisions, with the Commission d'accès à l'information as the complaint body. Every `qc` document carries a second, prominent review banner (counsel review, designated person, impact assessment, French-language requirement). Two optional config facts back it: `privacyPolicy.privacyOfficerTitle` and `privacyPolicy.impactAssessments` (the impact-assessment sentence renders only when attested). The "off until you choose" profiling assurance renders only under an opt-in consent model. **Not legal advice and not reviewed by counsel.**
|
|
54
|
+
|
|
55
|
+
CASL: when a form collects a newsletter sign-up, the `ca` **and** `qc` privacy policies now include a short "Commercial Electronic Messages" paragraph (consent, sender identification, unsubscribe). Existing `ca` documents change only for sites with such a form. No CSS change; no change to `ca` documents otherwise.
|
|
56
|
+
|
|
57
|
+
- 06d6112: Fix the EmDash `/design-system` admin page erroring on a tailwind site. It fetches `/vitops/design-manifest.json`, which the `tailwind` format never emitted and the integration only mirrored under `editor: true` (which also loads `editor.js` on every public page).
|
|
58
|
+
|
|
59
|
+
- **generator** (`tailwind` format only): `design-manifest.json` is now emitted alongside `tailwind.css`. No change to the CSS.
|
|
60
|
+
- **astro**: new `vitops({ editorManifest: true })` writes everything an outside host needs into `public/vitops/` (`design-manifest.json`, `editor.js`, a standalone `editor.css`, and `design-system-editor.html`) without loading the editor on public pages. `editor: true` still loads it from `<Head />`.
|
|
61
|
+
- **vite**: `editorManifestDir` also writes `vitops/editor.css`, from a throwaway `css`-format pass on tailwind sites.
|
|
62
|
+
- **emdash**: the `/design-system` admin page was unstyled and overflowed the admin, because the editor borrows the framework CSS and the admin never loads it. It now embeds `/vitops/design-system-editor.html` in an iframe, so the design system can't leak into the admin's styles. **Requires `editorManifest: true`** in `vitops({ … })`.
|
|
63
|
+
- **create**: the EmDash template sets `editorManifest: true`. Existing sites: add it to `vitops({ … })` in `astro.config.mjs`.
|
|
64
|
+
|
|
65
|
+
No consumer-facing change in the other toolchain packages.
|
|
66
|
+
|
|
67
|
+
- 6297337: `organization.address.streetAddress` is now optional (a home-based or service-area business with no public street address), and `organization.description` is a new optional, localisable field. The generated legal documents omit an absent street line; JSON-LD builders already skipped absent address fields. Non-breaking; no CSS change.
|
|
68
|
+
- 55061e2: **Added: `<SiteNav>` and `<SiteFooter>` for Astro/EmDash sites.** Both are tier-1 markup over the existing `sitenav` pattern (no runtime JS: `<details>` accordions + Popover API drawer). Non-breaking.
|
|
69
|
+
|
|
70
|
+
- `<SiteNav {...await loadNavMenus(getMenu)} />` renders EmDash menu `primary` (the default: depth 3, dropdowns from the `md` container width, drawer below). If a `primary-mobile` menu exists it replaces the nav below `md` (depth 1); if `primary-mega` exists it replaces it from `xl` (mega panels). Adding the menu in the admin is the whole change. Only one tier is in the accessibility tree at a time. An item's `titleAttr` is its description (EmDash has no description field yet).
|
|
71
|
+
- `<SiteFooter sitemap legal name>` renders `footer-sitemap` + `footer-legal` + a copyright line.
|
|
72
|
+
- `@getvitops/astro/components/sitenav.ts` exports the pure helpers (`loadNavMenus`, `selectTiers`, ...).
|
|
73
|
+
- CSS (`patterns/sitenav.css`, **all formats**): tier visibility classes (`.sitenav-tier--below-*`/`--from-*`), `.sitenav--mega`, `.sitenav__desc`, `.sitefooter*`, `aria-current` styling, reduced-motion and forced-colors handling. Existing `.sitenav` rules are unchanged, so the Bricks element is unaffected.
|
|
74
|
+
- `@getvitops/create` (emdash template): the layout now renders header and footer through these components, with the colour-scheme toggle in the header. It reads menus `primary`, `footer-sitemap` and `footer-legal`; a template whose seed still has the old `footer` menu renders no footer links until the seed is updated.
|
|
75
|
+
- `@getvitops/generator`: `TIERS.sitenav` now lists the two Astro components.
|
|
76
|
+
|
|
77
|
+
### Patch Changes
|
|
78
|
+
|
|
79
|
+
- Updated dependencies [161c7f3]
|
|
80
|
+
- Updated dependencies [175d16e]
|
|
81
|
+
- Updated dependencies [83d6f3c]
|
|
82
|
+
- Updated dependencies [b47df4d]
|
|
83
|
+
- @getvitops/utils@8.3.0-alpha.0
|
|
84
|
+
- @getvitops/config@8.3.0-alpha.0
|
|
85
|
+
|
|
86
|
+
## 8.2.0
|
|
87
|
+
|
|
88
|
+
### Minor Changes
|
|
89
|
+
|
|
90
|
+
- 65a13cc: **New: `site.experiments`** — runtime, per-visitor A/B content experiments for a page that
|
|
91
|
+
must stay at one URL (unlike `site.abTesting`'s existing build-time variant merge, which is
|
|
92
|
+
for a full design-system/CSS-level test via two manual deploys — `site.experiments` is the
|
|
93
|
+
opposite granularity: one page, one cookie, content branched in the page itself). Keyed by
|
|
94
|
+
an open experiment name, so more than one can run at once, each with its own consent
|
|
95
|
+
category, cookie name/lifetime, split ratio and pair of variant names.
|
|
96
|
+
|
|
97
|
+
**New `vitopsAbTesting()` middleware** (`@getvitops/astro/middleware`) — the first Astro
|
|
98
|
+
middleware this toolchain ships. Wire it into your own `src/middleware.ts`:
|
|
99
|
+
|
|
100
|
+
```ts
|
|
101
|
+
import { vitopsAbTesting } from '@getvitops/astro/middleware';
|
|
102
|
+
import config from '../site.json' with { type: 'json' };
|
|
103
|
+
|
|
104
|
+
export const onRequest = vitopsAbTesting({
|
|
105
|
+
experiments: config.site.experiments ?? {},
|
|
106
|
+
});
|
|
107
|
+
```
|
|
108
|
+
|
|
109
|
+
It assigns and cookie-pins a visitor to one of an experiment's two variants on first visit,
|
|
110
|
+
by weighted random (`splitRatio`) — and never touches an already-set cookie, so a recognized
|
|
111
|
+
query parameter named after the cookie itself (`?_ab_<key>=<variant>`) reliably forces that
|
|
112
|
+
variant, even overriding an existing assignment. Combined with `splitRatio: 0`, this is a
|
|
113
|
+
blue/green-style "100% on control, but a shared link reveals the other variant" mode with no
|
|
114
|
+
special-cased logic.
|
|
115
|
+
|
|
116
|
+
A page reads its own experiment's cookie (`Astro.cookies.get(cookieName)?.value`) and
|
|
117
|
+
renders whichever of its two hand-written blocks matches — there's no config-schema
|
|
118
|
+
representation of "the variant content," which is what keeps this lightweight regardless of
|
|
119
|
+
how many experiments are configured.
|
|
120
|
+
|
|
121
|
+
**`<Analytics />` gained a reporting hook**: each configured provider's bootstrap now
|
|
122
|
+
attaches an enabled experiment's assignment to the real pageview via that provider's own
|
|
123
|
+
mechanism — Plausible via `plausible.init({customProperties})`, Matomo via
|
|
124
|
+
`setCustomDimension` (needs a new `site.analytics.matomo.abVariantDimensions` entry — the
|
|
125
|
+
dimension must exist in your Matomo instance first), GA4 via a runtime-merged
|
|
126
|
+
`gtag('config', …)` call. A no-op when `site.experiments` is unset — zero behavior change
|
|
127
|
+
for every existing site.
|
|
128
|
+
|
|
129
|
+
Also new: `experimentDimensionWarnings(site)` (`@getvitops/generator`) — a build-time warning
|
|
130
|
+
(not a hard error) when an enabled experiment's site also configures Matomo but has no
|
|
131
|
+
matching `abVariantDimensions` entry.
|
|
132
|
+
|
|
133
|
+
**Not included in this pass, deliberately**: a mechanism for a fully prerendered
|
|
134
|
+
(`output: 'static'`) page — this needs a genuinely separate Cloudflare Worker Route + two
|
|
135
|
+
pre-built static variants, real but bounded follow-up work. Everything above applies to an
|
|
136
|
+
on-demand-rendered page (the current EmDash templates' default), which needs zero new
|
|
137
|
+
deploy infrastructure.
|
|
138
|
+
|
|
139
|
+
- 67e2088: **Config validation now rejects unknown keys instead of silently stripping them.** Every
|
|
140
|
+
object in `ConfigSchema` (the `Config` type — `organization`/`site` and everything nested under
|
|
141
|
+
them) switched from `z.object` to `z.strictObject`, matching what the published JSON Schema has
|
|
142
|
+
always asserted (`additionalProperties: false` at ~95 points) but the runtime validator never
|
|
143
|
+
actually enforced. A typo'd field (`site.ntoifications`, an add-on key on the wrong nesting
|
|
144
|
+
level, …) now fails `validateConfig()`/`vitops validate` with an `unrecognized_keys` error naming
|
|
145
|
+
the bad key, instead of validating successfully and doing nothing. `site.icons` stays
|
|
146
|
+
deliberately loose (`z.looseObject`, unchanged) — its per-set keys are open-ended by design.
|
|
147
|
+
|
|
148
|
+
**Breaking for any config carrying a field that isn't in the schema** — this is the point of the
|
|
149
|
+
change, but if a real config was relying on the old silent-strip behavior (an extra key kept
|
|
150
|
+
around for another tool to read), move it under a block the schema already permits `z.unknown()`
|
|
151
|
+
on (`site.cloudflare`, `site.tags`, `site.postTypes`) or open an issue.
|
|
152
|
+
|
|
153
|
+
Also:
|
|
154
|
+
|
|
155
|
+
- Added `configVersion` (optional, currently `1`) to `Config` — the document's own SHAPE
|
|
156
|
+
version, distinct from this package's release version (which is CSS-driven and bumps on every
|
|
157
|
+
design-system change regardless of `Config`'s shape). `defaultSiteConfig()` stamps it.
|
|
158
|
+
Exported as `CONFIG_VERSION`.
|
|
159
|
+
- Re-exported several `Config`-related types/values that were defined but missing from the
|
|
160
|
+
package root: `SiteReviews`, `SiteReviewLocation`, `SiteSearchConsoleSetup`,
|
|
161
|
+
`CrmProviderConfig`, `CRM_PROVIDERS`, `BIGIN_DATA_CENTERS`.
|
|
162
|
+
- Added a `@getvitops/generator/config` subpath export — `ConfigSchema`, `Config`,
|
|
163
|
+
`validateConfig`, `resolveConfig`, `isConfig`, `resolveInput` and friends, on their own, with no
|
|
164
|
+
`node:fs`/`node:crypto`/lightningcss in the import graph (verified: `publint` clean, and the
|
|
165
|
+
built chunk's own imports are `@zod/mini` + `@getvitops/utils`'s Worker-safe `ads`/`tracking`/
|
|
166
|
+
`color` subpaths only). This is what makes `Config` bundleable in a Cloudflare Worker — the
|
|
167
|
+
EmDash plugin's own admin route being the motivating case — without pulling in code that only
|
|
168
|
+
ever runs at build time.
|
|
169
|
+
|
|
170
|
+
- 89265af: **New package: `@getvitops/config`** — the site-config provider blocks (`site.crm`, `site.ads`,
|
|
171
|
+
`site.reviews`, `site.seo.indexing`, `site.searchConsole`, `site.notifications`, `site.tracking`,
|
|
172
|
+
`site.security`, `site.deployment`) as zod schemas and inferred types, with the closed provider lists
|
|
173
|
+
(`AD_PROVIDERS`, `CRM_PROVIDERS`, `BIGIN_DATA_CENTERS`) on a zod-free `./providers` entry. It joins the
|
|
174
|
+
`fixed` group.
|
|
175
|
+
|
|
176
|
+
Why: `@getvitops/utils` holds the provider code but could not import its configuration types — the schema
|
|
177
|
+
lived in `@getvitops/generator`, a layer above — so it kept a hand-copied mirror beside each module,
|
|
178
|
+
an adapter converting into it, and a test holding the copy to the schema. The mirrors are gone.
|
|
179
|
+
|
|
180
|
+
- **`@getvitops/generator`**: `AD_PROVIDERS`, `CRM_PROVIDERS`, `BIGIN_DATA_CENTERS`, `AdProvider` and
|
|
181
|
+
`CrmProviderConfig` are re-exported unchanged. No import path changes.
|
|
182
|
+
- **`@getvitops/utils`**: the config types in `./crm`, `./ads`, `./indexing`, `./onboarding`, `./reviews`,
|
|
183
|
+
`./tracking` and `./notify` (`CrmConfig`, `HubSpotConfig`, `BiginConfig`, `JobberConfig`,
|
|
184
|
+
`IndexNowConfig`, `SearchConsoleConfig`, `ReviewsConfig`, `ReviewLocationConfig`, `TrackingConfig`,
|
|
185
|
+
`NotificationsConfig`, `EmailChannelConfig`, `DomainSetup`, `AdPropertySetup`) are now inferred from the
|
|
186
|
+
schema. Same names, same fields, but **optional fields now follow zod's inference**, so a hand-written
|
|
187
|
+
object literal typed against them may need `| undefined` handled. `AdPropertySetup` gains the fields the
|
|
188
|
+
schema always had (`conversionActionId`, `reporting`).
|
|
189
|
+
- **`@getvitops/astro`**: `GetvitopsAdProperty` and `GetvitopsTrackingOptions` are aliases of the schema
|
|
190
|
+
types. `toCrmConfig()` and `toGoogleAdsConfig()` are **deprecated** — they returned their argument's
|
|
191
|
+
field unchanged; pass `config.site.crm` / `config.site.ads?.google` directly.
|
|
192
|
+
- **`@getvitops/cli`**: `vitops ads` now passes `conversionActionId` and `reporting` through to the planner;
|
|
193
|
+
it previously copied a hand-picked subset of the block and dropped them.
|
|
194
|
+
|
|
195
|
+
Formats reached: none — no CSS change.
|
|
196
|
+
|
|
197
|
+
- 0574388: Add `site.crm` — push form submissions into HubSpot, Zoho Bigin, or Jobber.
|
|
198
|
+
|
|
199
|
+
Client sites need form submissions to land in the client's CRM. The options available before this
|
|
200
|
+
— a HubSpot/Bigin/Jobber iframe or JS embed — fail on three counts: the markup can't be
|
|
201
|
+
white-labelled against this design system, nothing can operate on the submission before it leaves
|
|
202
|
+
the page, and the submission never touched `createConversionRoute()`, so `_ac` ad-click attribution
|
|
203
|
+
was lost and no conversion notification was planned.
|
|
204
|
+
|
|
205
|
+
This adds one server-side path for all three CRMs, hung off the conversion route that already
|
|
206
|
+
exists. `createConversionRoute()` gets a new `crm` option (`{ config, credentials, fetch?,
|
|
207
|
+
timeoutMs? }`); credentials arrive as a function of `locals`, exactly like the existing `binding`
|
|
208
|
+
option, and are never config. A failed CRM write never fails the request and is folded into the
|
|
209
|
+
notification email's new "Delivery" section, so a broken CRM token is never silent.
|
|
210
|
+
|
|
211
|
+
New `@getvitops/utils/crm` subpath: `deliverToCrm()`, `mapSubmission()`, `deriveLastName()` (Bigin
|
|
212
|
+
requires a last name; this derives one from an explicit field, a full name, or an email address,
|
|
213
|
+
refusing rather than fabricating one), `submitToHubSpot()`, `submitToBigin()`, `submitToJobber()`.
|
|
214
|
+
`identifySubmitter()` was promoted out of `notify`'s private name-guessing cascade into
|
|
215
|
+
`@getvitops/utils/tracking`, since the CRM mapper needed the whole identity (first/last name,
|
|
216
|
+
email, phone, company), not just a display name.
|
|
217
|
+
|
|
218
|
+
Jobber is GraphQL, not REST, and answers `200 OK` for failures — a rejected `clientCreate` returns
|
|
219
|
+
`userErrors[]`, a malformed request returns top-level `errors[]`. `submitToJobber()` classifies on
|
|
220
|
+
the response body, not just the status. Its refresh tokens also rotate by default: **Refresh Token
|
|
221
|
+
Rotation must be turned off** in the app's Developer Center settings, since a Worker reading a
|
|
222
|
+
static `VITOPS_JOBBER_REFRESH_TOKEN` cannot persist a rotated one — a rotated token is detected at
|
|
223
|
+
exchange time and reported with the fix, rather than surfacing an hour later as an opaque 401.
|
|
224
|
+
Attribution and unmapped extra fields render into a best-effort note by default (works on every
|
|
225
|
+
Jobber plan); a resubmitting visitor creates a second Jobber client unless `crm.jobber.dedupe: true`
|
|
226
|
+
is set once its lookup query has been verified against a live account (defaults `false` pending
|
|
227
|
+
that verification — see `jobber.ts`'s docblock).
|
|
228
|
+
|
|
229
|
+
New `site.crm` config block (`provider`, `hubspot`/`bigin`/`jobber`, `fieldMap`, `trackingFields`)
|
|
230
|
+
and two CLI commands: `vitops crm check` (fetches the live HubSpot property list or Bigin field
|
|
231
|
+
list and reports every mapped name that doesn't exist there — HubSpot rejects an unknown field
|
|
232
|
+
outright, Bigin silently drops it; Jobber has nothing live to check yet, since its mapped values
|
|
233
|
+
currently render as note text) and `vitops crm setup` (creates missing HubSpot properties; Bigin
|
|
234
|
+
and Jobber have no such API, so both always print a manual reminder).
|
|
235
|
+
|
|
236
|
+
The generated privacy policy now names HubSpot, Zoho Bigin, or Jobber when `site.crm` configures
|
|
237
|
+
one. Bigin gets the same cross-border-transfer derivation Matomo Cloud already gets: a
|
|
238
|
+
`dataCenter: 'ca'` asserts no transfer outside Canada, any other data centre does. Jobber's
|
|
239
|
+
operator (Octopusapp Inc. d/b/a Jobber) is Canadian, but its own privacy policy names the United
|
|
240
|
+
States as a processing location too, so the policy always asserts a cross-border transfer for
|
|
241
|
+
Jobber regardless of config — a Canadian operator is not read as Canadian-only storage.
|
|
242
|
+
|
|
243
|
+
Credentials: `VITOPS_HUBSPOT_ACCESS_TOKEN` for HubSpot; `VITOPS_ZOHO_CLIENT_ID` /
|
|
244
|
+
`VITOPS_ZOHO_CLIENT_SECRET` / `VITOPS_ZOHO_REFRESH_TOKEN` for Bigin; `VITOPS_JOBBER_CLIENT_ID` /
|
|
245
|
+
`VITOPS_JOBBER_CLIENT_SECRET` / `VITOPS_JOBBER_REFRESH_TOKEN` for Jobber. Never config.
|
|
246
|
+
|
|
247
|
+
No consumer-facing change in `@getvitops/astro` beyond the new `crm` route option and a build
|
|
248
|
+
warning when `site.crm` is configured with no route answering `TRACKING_ENDPOINT` — CSS/Bricks/
|
|
249
|
+
tailwind outputs are unaffected.
|
|
250
|
+
|
|
251
|
+
- c3c1bcb: **New: the design-system editor, mounted inside the EmDash admin.** `vitopsEmdash({
|
|
252
|
+
designSystemEditor: { config } })` (`config` is your `site.json`, e.g. `import config from
|
|
253
|
+
'./site.json' with { type: 'json' }`) adds a `/design-system` admin page mounting
|
|
254
|
+
`<wc-theme-editor>` — the same live token editor `@getvitops/vite`'s dev server has always
|
|
255
|
+
offered, now usable by the client directly, no dev server required. An edit is visible
|
|
256
|
+
site-wide immediately (a runtime `:root` override, served via a new plugin API route and
|
|
257
|
+
`page:fragments`) and — once `site.deployment.repo`/`branch` and a `VITOPS_GITHUB_TOKEN`
|
|
258
|
+
secret are set (see the scaffold's `SETUP.html`) — auto-commits to your `dev` branch, so it
|
|
259
|
+
reaches the git-tracked `site.json` through your repo's normal review path rather than
|
|
260
|
+
becoming a second, unreviewed source of truth. Until that's configured, the live half still
|
|
261
|
+
works; only the auto-commit half is skipped (and says so in the admin page). The save route
|
|
262
|
+
runs the exact same accessibility (APCA contrast) gate `generate()` enforces at build time,
|
|
263
|
+
synchronously, before persisting — nothing non-compliant ever goes live even momentarily.
|
|
264
|
+
|
|
265
|
+
Requires `@astrojs/react` + `react` + `react-dom` (already scaffold dependencies, for
|
|
266
|
+
EmDash's own admin) and `@getvitops/core` (added as a new scaffold dependency this release).
|
|
267
|
+
Declares two new plugin capabilities when enabled: `network:request` (for the GitHub
|
|
268
|
+
auto-commit) and `hooks.page-fragments:register` (for the live override) — reviewed once at
|
|
269
|
+
plugin-install time.
|
|
270
|
+
|
|
271
|
+
**Breaking for `@getvitops/emdash`**: `vitopsEmdash()`'s `hooks.page-fragments:register`
|
|
272
|
+
capability is now also granted whenever `designSystemEditor` is set, independent of the
|
|
273
|
+
existing `scripts: 'fragments'` option (which governs something unrelated — web-component
|
|
274
|
+
script delivery). No change if you don't use `designSystemEditor`.
|
|
275
|
+
|
|
276
|
+
Also new, supporting the above:
|
|
277
|
+
|
|
278
|
+
- `@getvitops/generator`: `validateColorContract()` and `buildReverseIndex()` (both
|
|
279
|
+
`/tokens` subpath, Worker-safe) — the monotonicity/contrast gate and the
|
|
280
|
+
`design-manifest.json` reverse-index, each extracted from `generate()`'s `build()` into
|
|
281
|
+
standalone, non-throwing functions. `patchToOverrideCss()` turns a design-system patch
|
|
282
|
+
into the `:root` override `<style>` block. `designSystemPath`/`getAt`/`setAt`/
|
|
283
|
+
`mergePatch`/`formatConfig` moved from `@getvitops/vite` into `@getvitops/generator/config`
|
|
284
|
+
(re-exported from `@getvitops/vite` unchanged) — one "patch a config file's design-system
|
|
285
|
+
subtree" implementation shared by the dev endpoint and the new plugin route. New
|
|
286
|
+
`site.deployment.repo`/`branch` config fields.
|
|
287
|
+
- `@getvitops/utils`: new `/git` subpath — `commitFile()`, a Worker-safe (fetch-only)
|
|
288
|
+
wrapper over GitHub's Contents API.
|
|
289
|
+
- `@getvitops/core`: `<wc-theme-editor>` gains a `save-endpoint` attribute (default
|
|
290
|
+
unchanged — `/__vitops/design-system`) so the same element can target the new plugin
|
|
291
|
+
route instead of the dev-only Vite endpoint.
|
|
292
|
+
|
|
293
|
+
- 0574388: Add `vitops email setup|check` — connect `site.notifications.email.from` to whether its domain is
|
|
294
|
+
actually onboarded onto an outbound email provider.
|
|
295
|
+
|
|
296
|
+
Before this, nothing checked that. `EMAIL.send()` throws `E_SENDER_NOT_VERIFIED` at runtime with no
|
|
297
|
+
build or deploy-time warning — the failure mode is silent and production-only.
|
|
298
|
+
|
|
299
|
+
`vitops email setup [domain]` onboards a sending domain (currently Cloudflare Email Sending only)
|
|
300
|
+
and polls DNS until its SPF/DKIM/DMARC records resolve or a timeout is reached; `vitops email
|
|
301
|
+
check` is the non-blocking deploy guard both `@getvitops/create` templates' `deploy`/`deploy:dev`
|
|
302
|
+
scripts now run first (`--strict` makes it a real gate; opt out per-site with
|
|
303
|
+
`site.notifications.email.verifySender: false`, or globally with `--skip` /
|
|
304
|
+
`VITOPS_SKIP_EMAIL_CHECK`).
|
|
305
|
+
|
|
306
|
+
Two things worth knowing before running it:
|
|
307
|
+
|
|
308
|
+
- **The onboarded entity is an exact name, and it is usually a subdomain**
|
|
309
|
+
(`send.yourdomain.com`), never the apex your `from` address names. When the exact domain isn't
|
|
310
|
+
onboarded but a sibling in the same zone already is, the command reports that as the fix — send
|
|
311
|
+
from the onboarded subdomain — rather than silently trying to onboard the apex, which rewrites
|
|
312
|
+
`_dmarc.<domain>` and can tighten DMARC policy on every other mail your domain sends.
|
|
313
|
+
- Candidates come from `site.notifications.email.from` (or its derived `noreply@<domains
|
|
314
|
+
.canonical>` default) first; a best-effort scan of `--src` only supplements it, and never guesses
|
|
315
|
+
at a computed `from` value.
|
|
316
|
+
|
|
317
|
+
New `@getvitops/utils/email` subpath and `site.notifications.email.verifySender` config field.
|
|
318
|
+
Needs `CLOUDFLARE_API_TOKEN` with an Email Sending scope in addition to the `Zone:Read` other
|
|
319
|
+
commands already use.
|
|
320
|
+
|
|
321
|
+
- 1a8231d: **EmDash blocks are now a tier of the component manifest, and there are four more of them.**
|
|
322
|
+
|
|
323
|
+
`@getvitops/emdash` adds `vitops.gallery`, `vitops.cards`, `vitops.accordion` and `vitops.marquee`
|
|
324
|
+
(each a thin adapter over the matching `@getvitops/astro` component), and `vitops.carousel` now edits
|
|
325
|
+
its slides as a **Slides repeater** instead of two newline-separated fields. Blocks saved before this
|
|
326
|
+
keep rendering: with no `slides`, the carousel still reads `images`/`alts`, and both legacy fields stay
|
|
327
|
+
in the editor, labelled as legacy. The editor's field type now admits `repeater` and `media_picker`.
|
|
328
|
+
|
|
329
|
+
- **New `blocks` plugin option** — `vitopsEmdash({ blocks: { include, exclude, extra } })`. `include` is
|
|
330
|
+
an allowlist, `exclude` drops blocks, `extra` adds the site's own **editor** definitions (the site
|
|
331
|
+
supplies their renderers via `<PortableText components={{ type }}>`, the same route that overrides a
|
|
332
|
+
`vitops.*` renderer). An unknown name or a colliding `type` throws at plugin construction.
|
|
333
|
+
- Link and image URLs in blocks with a scheme other than `http`, `https`, `mailto` or `tel` are
|
|
334
|
+
dropped (see the security entry).
|
|
335
|
+
|
|
336
|
+
`@getvitops/generator`: `TierEntry` gains an optional `emdash: string[]` (block `_type`s) and `Tier`
|
|
337
|
+
gains `'emdash'`, so `tierPatterns('emdash')` projects it. **Breaking only for code that exhaustively
|
|
338
|
+
switches on `Tier` or builds a `Record<Tier, …>`** — add an `emdash` case. The agent doc
|
|
339
|
+
`vitops docs components` gains an "EmDash blocks" table and the docsite a `components/emdash` page. The
|
|
340
|
+
guard that every block and every `emdash` entry agree lives in `@getvitops/emdash`'s `blocks.test.ts`,
|
|
341
|
+
because the generator cannot import the plugin.
|
|
342
|
+
|
|
343
|
+
`@getvitops/create`: the EmDash template's agent guidance explains the blocks, their limits (a block
|
|
344
|
+
cannot contain blocks; `vitops.*` blocks do not render inside EmDash's built-in `columns`) and the
|
|
345
|
+
`blocks` option.
|
|
346
|
+
|
|
347
|
+
Formats reached: none — no CSS change. Other packages in the `fixed` group: No consumer-facing change
|
|
348
|
+
in this package.
|
|
349
|
+
|
|
350
|
+
- 8d8aad7: **New: server-side Google Ads offline conversion upload.** Set
|
|
351
|
+
`site.ads.google.reporting: "server"` and a matching
|
|
352
|
+
`conversionActionId` (a numeric id or full
|
|
353
|
+
`customers/…/conversionActions/…` resource name — **not** the same value
|
|
354
|
+
as `conversionLabel`, which is the opaque label the web tag uses and
|
|
355
|
+
which the Ads API's upload endpoint doesn't accept) and
|
|
356
|
+
`createConversionRoute`'s new `googleAdsOffline` option uploads the
|
|
357
|
+
conversion via `ConversionUploadService.UploadClickConversions` instead
|
|
358
|
+
of relying on the visitor's browser to fire `gtag.js`. An action with no
|
|
359
|
+
matching `conversionActionId` keeps firing client-side regardless of
|
|
360
|
+
`reporting`, so turning it on before the id is filled in never silently
|
|
361
|
+
drops a conversion. `@getvitops/astro/routes` exports a matching
|
|
362
|
+
`toGoogleAdsConfig(config)` adapter, alongside `toCrmConfig`.
|
|
363
|
+
|
|
364
|
+
Structurally typed against Google's documented REST shape, not verified
|
|
365
|
+
against a live Ads account — no developer-token/MCC access exists
|
|
366
|
+
anywhere in this toolchain's own setup to test against. Wrapped so a
|
|
367
|
+
wrong guess degrades to a stated failure reason (folded into the
|
|
368
|
+
conversion email's "Delivery" section, same as a failed CRM write) rather
|
|
369
|
+
than breaking, the same caveat `@getvitops/utils/crm`'s Jobber executor
|
|
370
|
+
already carries for its own unverified surfaces.
|
|
371
|
+
|
|
372
|
+
`vitops ads lint` now flags a `reporting: "server"` property whose
|
|
373
|
+
configured actions have no matching `conversionActionId` — not a
|
|
374
|
+
failure, since that action keeps firing client-side, but worth surfacing
|
|
375
|
+
rather than leaving someone to wonder why nothing changed.
|
|
376
|
+
|
|
377
|
+
Not built: `vitops ads provision` (upserting conversion actions via
|
|
378
|
+
`ConversionActionService` and printing back both the label and the
|
|
379
|
+
resource id) — same credential blocker, and deliberately left unscoped
|
|
380
|
+
until there's a live account to verify the request/response shape
|
|
381
|
+
against.
|
|
382
|
+
|
|
383
|
+
- dde3df3: **`site.ads.google.conversionLabel` now accepts a per-action map** —
|
|
384
|
+
`{ call?, email?, form? }` — instead of one bare string shared by every
|
|
385
|
+
interaction. A bare string still works (back-compat, treated as `call`), but a
|
|
386
|
+
call and an email click used to report as the **same** Google Ads conversion
|
|
387
|
+
action, indistinguishable in the account; each action now gets its own
|
|
388
|
+
`send_to` target.
|
|
389
|
+
|
|
390
|
+
**New: `vitopsFormConversion(formEl)`**, exposed as `window.vitopsFormConversion`
|
|
391
|
+
by `<Tracking />`'s capture script — call it from your own fetch-based form
|
|
392
|
+
submit handler once the POST to `createConversionRoute()` succeeds, to fire the
|
|
393
|
+
`form` Google Ads conversion action client-side (mirroring the existing
|
|
394
|
+
tel:/mailto: click handler, which now fires `call`/`email` instead of sharing
|
|
395
|
+
one label). Idempotent per form element. There's deliberately no server-side
|
|
396
|
+
equivalent for a no-JS `<form method="post">` submission — that's a full-page
|
|
397
|
+
redirect with no script to fire from, and a real server-side Google Ads
|
|
398
|
+
conversion needs the Ads API's click-conversion upload (developer token + MCC
|
|
399
|
+
access), tracked separately in the vNext backlog.
|
|
400
|
+
|
|
401
|
+
**Fixed: the conversion route's rate limiter no longer trusts the client's
|
|
402
|
+
raw `event` field.** It used to key on whatever string a client sent
|
|
403
|
+
(`` `${ip}:${body.event}` ``), so varying that value on every request let a
|
|
404
|
+
form submission dodge the per-IP rate limit entirely — only `'call'`/`'email'`
|
|
405
|
+
were ever recognized, and anything else got its own fresh bucket. The key now
|
|
406
|
+
uses the server's own classification (`call` | `email` | `form`), closing that
|
|
407
|
+
bypass without changing the existing call-vs-email-don't-compete behavior.
|
|
408
|
+
|
|
409
|
+
**New: `site.notifications.actionUrl`** — an optional action-link template
|
|
410
|
+
added to the conversion-notification email as an "Actions" section.
|
|
411
|
+
`{clickId}` is replaced with the ad click ID captured on `_ac` (gclid, fbclid,
|
|
412
|
+
etc.); the section is omitted when the event has no click ID to interpolate,
|
|
413
|
+
rather than emailing a broken link.
|
|
414
|
+
|
|
415
|
+
- d9f213b: **New: `site.plan`** — records a commercial plan/add-on entitlement (`{ key, addons? }`),
|
|
416
|
+
distinct from a section's own `enabled` flag: `enabled` answers "does this site do X at all"
|
|
417
|
+
(a capability switch); `plan` answers "is this client entitled to turn X on" (a commercial
|
|
418
|
+
switch one layer above). Absence means unrestricted, matching every other opt-in block in
|
|
419
|
+
`Config`.
|
|
420
|
+
|
|
421
|
+
`PLAN_CAPABILITIES` (new export) is the small, hand-authored table of which add-on gates
|
|
422
|
+
which `Config` field — today: `ab_testing` gates `site.abTesting.enabled`, any
|
|
423
|
+
`environments.*.variant`, **and any enabled `site.experiments` entry** (that last one was
|
|
424
|
+
missed in this feature's first pass and fixed before release — found while planning this
|
|
425
|
+
toolchain's downstream adoption, not by a consumer hitting it), `server_side_tracking` gates
|
|
426
|
+
`site.tracking.enabled`. `validateConfig` now rejects a config that turns on a gated feature
|
|
427
|
+
without listing the owning add-on in `site.plan.addons`, naming the add-on in the error.
|
|
428
|
+
|
|
429
|
+
**Breaking**: an existing config using `site.abTesting.enabled`/`environments.*.variant` or
|
|
430
|
+
`site.tracking.enabled` now needs a matching `site.plan.addons` entry, or `validateConfig`
|
|
431
|
+
rejects it. Add `site.plan: { key: '<your-plan>', addons: ['ab_testing'] }` (and/or
|
|
432
|
+
`'server_side_tracking'`) to a config already using either feature.
|
|
433
|
+
|
|
434
|
+
- 7fd5f8f: **`site.seo` is now the single source of site-level SEO defaults** (ADR-0002). Until now `site.seo`
|
|
435
|
+
declared `titleTemplate`, verification tokens, `openGraph` and `twitter` that nothing read, while
|
|
436
|
+
`<Seo />` took a second, differently-shaped vocabulary through `vitops({ seo })` that no other stack
|
|
437
|
+
could see.
|
|
438
|
+
|
|
439
|
+
- `SeoSchema` moves to `@getvitops/config` and gains what only the integration had: `siteName`,
|
|
440
|
+
`defaultTitle`, `defaultDescription`, `openGraph.localeAlternates` and `openGraph.image` as
|
|
441
|
+
`{ src, alt, width, height, type }`. Text fields are localisable.
|
|
442
|
+
- `@getvitops/astro` exports `resolveSeoDefaults(config, { locale, env })`. The integration reads
|
|
443
|
+
`site.seo` from the site config it already loads; `siteName` falls back to `organization.name`, and
|
|
444
|
+
the environment's `robots` wins over `site.seo.robots`, the same fact `vitops robots` and
|
|
445
|
+
`search notify` read.
|
|
446
|
+
- `vitops({ seo })` keeps working unchanged and is now an **override** merged over the config, field by
|
|
447
|
+
field (`openGraph` and `twitter` merge key by key).
|
|
448
|
+
- **Removed from the config schema** (nothing read them): `descriptionTemplate` → use
|
|
449
|
+
`defaultDescription`; `openGraph.siteName`/`title`/`description` and
|
|
450
|
+
`twitter.title`/`description`/`image` → set per page on `<Seo />`, or the site-wide values above.
|
|
451
|
+
A config that still carries one now fails `vitops validate` with an unrecognised-key error.
|
|
452
|
+
- `openGraph.type` is now `website | article | profile`.
|
|
453
|
+
|
|
454
|
+
### Patch Changes
|
|
455
|
+
|
|
456
|
+
- 779d22e: Add a `vitops.hero` EmDash block (eyebrow, heading, text, two links, optional image; centered or split layout) built on the `.centered`/`.split` layout primitives. The EmDash template's home page now seeds one of each of hero, cards, accordion and marquee. `@getvitops/generator`: the `split` pattern lists the new block in `TIERS`.
|
|
457
|
+
- 0574388: Bump `emdash`/`@emdash-cms/cloudflare` peers to `>=1.0.0 <2` and `vite-plus` to `^1.0.0`, following
|
|
458
|
+
both projects' GA releases.
|
|
459
|
+
|
|
460
|
+
**Breaking for `@getvitops/emdash` consumers on EmDash 0.x**: the peer range no longer admits
|
|
461
|
+
`emdash@0.31.x`/`@emdash-cms/cloudflare@0.31.x`. Upgrade `emdash` and `@emdash-cms/cloudflare` to
|
|
462
|
+
`^1.0.1` first — nothing in `@getvitops/emdash`'s own plugin surface (`admin.portableTextBlocks`,
|
|
463
|
+
the `hooks.page-fragments:register` capability, `definePlugin()`) changed against EmDash 1.0, so no
|
|
464
|
+
code here needed to change, only the declared range.
|
|
465
|
+
|
|
466
|
+
The `emdash` scaffold template (`vp create @getvitops:emdash`) picks up several fixes alongside
|
|
467
|
+
the version bump:
|
|
468
|
+
|
|
469
|
+
- `tailwindcss`/`@tailwindcss/vite` are now real dependencies — the template selected
|
|
470
|
+
`css.format: 'tailwind'` without installing either, which threw on build.
|
|
471
|
+
- `site.json` now sets `site.icons.sprite: true` plus the full `semantic` list the
|
|
472
|
+
`vitops.actionLink` block's icon picker offers — previously its sprite icons pointed at a file
|
|
473
|
+
the build never generated.
|
|
474
|
+
- `astro.config.mjs` sets `migrations: { runtime: 'check', dev: 'auto' }`, and
|
|
475
|
+
`site-deploy-emdash.yml` runs EmDash 1.0's documented `build → status → apply → deploy → check`
|
|
476
|
+
migration sequence explicitly, rather than relying solely on first-request auto-migration. This
|
|
477
|
+
needs a new GitHub Environment variable, `EMDASH_TARGET_FINGERPRINT` (one per environment),
|
|
478
|
+
which only a human can set — run `emdash migrate --status` against the real deployed database,
|
|
479
|
+
review the printed identity, then `gh variable set EMDASH_TARGET_FINGERPRINT --env <env> --body
|
|
480
|
+
<value>`. Deploys fail clearly on the migration-apply step until it's set.
|
|
481
|
+
- `lint.options.typeAware`/`typeCheck` in the template's own `vite.config.ts` are commented out,
|
|
482
|
+
not deleted: vite-plus 1.0 only honours that option in a project's actual root config, and hard
|
|
483
|
+
errors on it in a nested one (which this file is, sitting inside the `design-system` monorepo).
|
|
484
|
+
Uncomment it once scaffolded into its own repo.
|
|
485
|
+
|
|
486
|
+
No config schema changes. `@getvitops/generator` also switches from the `zod/mini` subpath import
|
|
487
|
+
to the standalone `@zod/mini` package (zod 4.5+; same API) and fixes `validateConfig()` silently
|
|
488
|
+
losing the offending key name on an `unrecognized_keys` error (e.g. a typo'd `site.ads` platform
|
|
489
|
+
now names the bad key instead of reporting a bare "Invalid input").
|
|
490
|
+
|
|
491
|
+
Also fixed: `vp check`'s `tsconfig-error` for `packages/create/templates/emdash/tsconfig.json`
|
|
492
|
+
(`extends: "astro/tsconfigs/strict"`, unresolvable from inside this monorepo since the template
|
|
493
|
+
has no `node_modules` of its own — a pre-existing, structural gap, newly surfaced as a hard error
|
|
494
|
+
by oxlint 1.85's deeper config discovery, and blocking on every commit since this repo's pre-commit
|
|
495
|
+
hook runs `vp check --fix`). Fixed with a new `scripts/link-template-deps.mjs`, wired as the root
|
|
496
|
+
`postinstall` script: symlinks `astro`/`emdash`/`@emdash-cms/cloudflare` into the template's own
|
|
497
|
+
`node_modules/` (resolved fresh from sibling packages that already depend on them for real, not a
|
|
498
|
+
hardcoded `.pnpm` hash) and writes a minimal `astro:content` ambient shim at `.astro/types.d.ts`
|
|
499
|
+
(the same path `astro sync` would generate for real). Both are gitignored and only ever exist
|
|
500
|
+
inside this monorepo's copy of the template — a real scaffold installs its own real copies instead.
|
|
501
|
+
Considered and rejected first: inlining astro's strict preset (drift risk against astro's own
|
|
502
|
+
evolving default) and making the template a real pnpm workspace member (would change how its
|
|
503
|
+
`@getvitops/*: "latest"` deps resolve locally, which `templates.test.ts` exists to protect).
|
|
504
|
+
|
|
505
|
+
**Known gap, not fixed here**: `packages/astro/src/components/container.test.ts` (the
|
|
506
|
+
`AstroContainer`-based render tests) fails to even load under vite-plus 1.0's Vite 8 — direct
|
|
507
|
+
`.astro` imports in a Vitest file need the Astro Vite plugin registered, which nothing in this
|
|
508
|
+
repo's test config actually does (Astro's own docs call for `getViteConfig()`, which doesn't
|
|
509
|
+
compose with vite-plus's own `defineConfig`). A type-level shim
|
|
510
|
+
(`packages/astro/src/components/astro-modules.d.ts`) fixes `vp check`'s type-aware pass, but the
|
|
511
|
+
runtime `vp test` failure for this one file is still open.
|
|
512
|
+
|
|
513
|
+
- 1a3e811: **Fixed: `<Icon />` no longer crashes a Cloudflare Workers SSR page.** It statically imported
|
|
514
|
+
`loadIconSvg`/`spriteId` from `@getvitops/generator`'s root entry, which also re-exports
|
|
515
|
+
`generate.ts` — reaching lightningcss, whose native binary Workers cannot `require()`.
|
|
516
|
+
Rendering `<Icon>` anywhere in an SSR project 500'd every request, regardless of
|
|
517
|
+
`icons.engine`, since the import itself (not just calling the function) pulled lightningcss
|
|
518
|
+
into the bundle. Fixed by publishing a new Worker-safe `@getvitops/generator/icons-sprite`
|
|
519
|
+
subpath (`buildIconSprite`/`loadIconSvg`/`spriteId`, no lightningcss in its import graph) and
|
|
520
|
+
switching `<Icon />` to import from there.
|
|
521
|
+
- Updated dependencies [89265af]
|
|
522
|
+
- Updated dependencies [0574388]
|
|
523
|
+
- Updated dependencies [c3c1bcb]
|
|
524
|
+
- Updated dependencies [0574388]
|
|
525
|
+
- Updated dependencies [0574388]
|
|
526
|
+
- Updated dependencies [8d8aad7]
|
|
527
|
+
- Updated dependencies [0574388]
|
|
528
|
+
- Updated dependencies [0574388]
|
|
529
|
+
- Updated dependencies [dde3df3]
|
|
530
|
+
- Updated dependencies [7f4f50e]
|
|
531
|
+
- Updated dependencies [7fd5f8f]
|
|
532
|
+
- Updated dependencies [ceb8310]
|
|
533
|
+
- @getvitops/config@8.2.0
|
|
534
|
+
- @getvitops/utils@8.2.0
|
|
535
|
+
|
|
3
536
|
## 8.2.0-alpha.0
|
|
4
537
|
|
|
5
538
|
### Minor Changes
|
|
@@ -93,7 +626,7 @@
|
|
|
93
626
|
|
|
94
627
|
This adds one server-side path for all three CRMs, hung off the conversion route that already
|
|
95
628
|
exists. `createConversionRoute()` gets a new `crm` option (`{ config, credentials, fetch?,
|
|
96
|
-
timeoutMs? }`); credentials arrive as a function of `locals`, exactly like the existing `binding`
|
|
629
|
+
timeoutMs? }`); credentials arrive as a function of `locals`, exactly like the existing `binding`
|
|
97
630
|
option, and are never config. A failed CRM write never fails the request and is folded into the
|
|
98
631
|
notification email's new "Delivery" section, so a broken CRM token is never silent.
|
|
99
632
|
|
|
@@ -138,8 +671,8 @@ timeoutMs? }`); credentials arrive as a function of `locals`, exactly like the e
|
|
|
138
671
|
tailwind outputs are unaffected.
|
|
139
672
|
|
|
140
673
|
- c3c1bcb: **New: the design-system editor, mounted inside the EmDash admin.** `vitopsEmdash({
|
|
141
|
-
designSystemEditor: { config } })` (`config` is your `site.json`, e.g. `import config from
|
|
142
|
-
'./site.json' with { type: 'json' }`) adds a `/design-system` admin page mounting
|
|
674
|
+
designSystemEditor: { config } })` (`config` is your `site.json`, e.g. `import config from
|
|
675
|
+
'./site.json' with { type: 'json' }`) adds a `/design-system` admin page mounting
|
|
143
676
|
`<wc-theme-editor>` — the same live token editor `@getvitops/vite`'s dev server has always
|
|
144
677
|
offered, now usable by the client directly, no dev server required. An edit is visible
|
|
145
678
|
site-wide immediately (a runtime `:root` override, served via a new plugin API route and
|
|
@@ -186,7 +719,7 @@ designSystemEditor: { config } })` (`config` is your `site.json`, e.g. `import c
|
|
|
186
719
|
|
|
187
720
|
`vitops email setup [domain]` onboards a sending domain (currently Cloudflare Email Sending only)
|
|
188
721
|
and polls DNS until its SPF/DKIM/DMARC records resolve or a timeout is reached; `vitops email
|
|
189
|
-
check` is the non-blocking deploy guard both `@getvitops/create` templates' `deploy`/`deploy:dev`
|
|
722
|
+
check` is the non-blocking deploy guard both `@getvitops/create` templates' `deploy`/`deploy:dev`
|
|
190
723
|
scripts now run first (`--strict` makes it a real gate; opt out per-site with
|
|
191
724
|
`site.notifications.email.verifySender: false`, or globally with `--skip` /
|
|
192
725
|
`VITOPS_SKIP_EMAIL_CHECK`).
|
|
@@ -198,7 +731,7 @@ check` is the non-blocking deploy guard both `@getvitops/create` templates' `dep
|
|
|
198
731
|
from the onboarded subdomain — rather than silently trying to onboard the apex, which rewrites
|
|
199
732
|
`_dmarc.<domain>` and can tighten DMARC policy on every other mail your domain sends.
|
|
200
733
|
- Candidates come from `site.notifications.email.from` (or its derived `noreply@<domains
|
|
201
|
-
.canonical>` default) first; a best-effort scan of `--src` only supplements it, and never guesses
|
|
734
|
+
.canonical>` default) first; a best-effort scan of `--src` only supplements it, and never guesses
|
|
202
735
|
at a computed `from` value.
|
|
203
736
|
|
|
204
737
|
New `@getvitops/utils/email` subpath and `site.notifications.email.verifySender` config field.
|
|
@@ -280,7 +813,7 @@ check` is the non-blocking deploy guard both `@getvitops/create` templates' `dep
|
|
|
280
813
|
needs a new GitHub Environment variable, `EMDASH_TARGET_FINGERPRINT` (one per environment),
|
|
281
814
|
which only a human can set — run `emdash migrate --status` against the real deployed database,
|
|
282
815
|
review the printed identity, then `gh variable set EMDASH_TARGET_FINGERPRINT --env <env> --body
|
|
283
|
-
<value>`. Deploys fail clearly on the migration-apply step until it's set.
|
|
816
|
+
<value>`. Deploys fail clearly on the migration-apply step until it's set.
|
|
284
817
|
- `lint.options.typeAware`/`typeCheck` in the template's own `vite.config.ts` are commented out,
|
|
285
818
|
not deleted: vite-plus 1.0 only honours that option in a project's actual root config, and hard
|
|
286
819
|
errors on it in a nested one (which this file is, sitting inside the `design-system` monorepo).
|
|
@@ -623,7 +1156,7 @@ check` is the non-blocking deploy guard both `@getvitops/create` templates' `dep
|
|
|
623
1156
|
It is **required for a user credential and wrong for a service account**, which is why it is not
|
|
624
1157
|
sent unconditionally. A user credential authenticates through a shared OAuth client that owns no
|
|
625
1158
|
project, so Google refuses the call outright — measured: `403 "requires a quota project, which is
|
|
626
|
-
not set by default"`. A service account already belongs to a project, and pointing it at a
|
|
1159
|
+
not set by default"`. A service account already belongs to a project, and pointing it at a
|
|
627
1160
|
different one would newly require `serviceusage.services.use` there, turning a working CI
|
|
628
1161
|
credential into a 403 for no benefit.
|
|
629
1162
|
|
|
@@ -648,7 +1181,7 @@ not set by default"`. A service account already belongs to a project, and pointi
|
|
|
648
1181
|
**Fixed: an ADC file in `GOOGLE_APPLICATION_CREDENTIALS` no longer hard-exits.** That variable is
|
|
649
1182
|
Google's own convention _and_ where gcloud writes ADC, so the same variable carries two kinds of
|
|
650
1183
|
credential. Parsing an `authorized_user` file as a service account failed with `service account
|
|
651
|
-
JSON is missing client_email or private_key` — while holding a credential the command already knew
|
|
1184
|
+
JSON is missing client_email or private_key` — while holding a credential the command already knew
|
|
652
1185
|
how to use. It is now discriminated on `type`, the field Google's own libraries switch on.
|
|
653
1186
|
|
|
654
1187
|
**Also fixed — `search setup --dry` and `--check` are not offline.** The docs implied otherwise. The
|
|
@@ -729,7 +1262,7 @@ JSON is missing client_email or private_key` — while holding a credential the
|
|
|
729
1262
|
created, with `patterns/counter.css`'s own reduced-motion query as a second belt.
|
|
730
1263
|
- The accessible name stays the final value for the whole animation: `.counter__value` is toggled
|
|
731
1264
|
to the framework's `.sr-only` utility while running, never hidden — `display: none`/`visibility:
|
|
732
|
-
hidden` would pull it out of the accessible-name computation along with the visual hiding, and a
|
|
1265
|
+
hidden` would pull it out of the accessible-name computation along with the visual hiding, and a
|
|
733
1266
|
screen reader must never be read ~60 intermediate numbers. The animated text lives in a separate
|
|
734
1267
|
`aria-hidden` presentation span it never reaches.
|
|
735
1268
|
- `data-from` (default `0`), `data-duration` (ms, default `1200`), `data-easing` (`linear` |
|
|
@@ -1488,28 +2021,20 @@ hidden` would pull it out of the accessible-name computation along with the visu
|
|
|
1488
2021
|
{
|
|
1489
2022
|
"designSystem": {
|
|
1490
2023
|
"themes": {
|
|
1491
|
-
"default": {
|
|
1492
|
-
/* … */
|
|
1493
|
-
},
|
|
2024
|
+
"default": {/* … */},
|
|
1494
2025
|
},
|
|
1495
2026
|
},
|
|
1496
2027
|
"organization": {
|
|
1497
2028
|
"name": "Acme",
|
|
1498
2029
|
"contact": "hq",
|
|
1499
2030
|
"locations": {
|
|
1500
|
-
"hq": {
|
|
1501
|
-
/* … */
|
|
1502
|
-
},
|
|
2031
|
+
"hq": {/* … */},
|
|
1503
2032
|
},
|
|
1504
2033
|
},
|
|
1505
2034
|
"site": {
|
|
1506
2035
|
"defaultLocale": "en",
|
|
1507
|
-
"domains": {
|
|
1508
|
-
|
|
1509
|
-
},
|
|
1510
|
-
"analytics": {
|
|
1511
|
-
/* … */
|
|
1512
|
-
},
|
|
2036
|
+
"domains": {/* … */},
|
|
2037
|
+
"analytics": {/* … */},
|
|
1513
2038
|
},
|
|
1514
2039
|
}
|
|
1515
2040
|
```
|
|
@@ -1603,7 +2128,7 @@ hidden` would pull it out of the accessible-name computation along with the visu
|
|
|
1603
2128
|
`animate-view` / `animate-scroll` / `animate-trigger` / `transition`, and every effect
|
|
1604
2129
|
gained an `open-<fx>` state variant. Overlays now state _where they start_
|
|
1605
2130
|
(`--translate-x-from`) instead of owning a keyframe: `class="drawer drawer--right
|
|
1606
|
-
open-fade-in"` composes slide and fade. Applied at zero specificity with identity
|
|
2131
|
+
open-fade-in"` composes slide and fade. Applied at zero specificity with identity
|
|
1607
2132
|
defaults, so a popover that sets no effect vars is unchanged.
|
|
1608
2133
|
- **One scrim.** `--scrim` / `--scrim-filter` tokens and `.no-scrim` in `popover.css`,
|
|
1609
2134
|
replacing 18 `::backdrop` blocks across seven partials. `.drawer--modeless` is aliased.
|
|
@@ -1918,8 +2443,8 @@ open-fade-in"` composes slide and fade. Applied at zero specificity with identit
|
|
|
1918
2443
|
replace it with `--anim-range`. `--stagger-range-step` is now a viewport length (`5vh`) to match.
|
|
1919
2444
|
|
|
1920
2445
|
**Fixed: every journey ran on a truncated range.** The generator emitted `animation-range: entry
|
|
1921
|
-
exit`; lightningcss parsed the end of that shorthand as `exit 0%` rather than the spec's `exit
|
|
1922
|
-
100%`, so the bundle shipped `entry exit 0%`. Journeys reach their `100%` keyframe — the hidden
|
|
2446
|
+
exit`; lightningcss parsed the end of that shorthand as `exit 0%` rather than the spec's `exit
|
|
2447
|
+
100%`, so the bundle shipped `entry exit 0%`. Journeys reach their `100%` keyframe — the hidden
|
|
1923
2448
|
`from` state — as the element hits the top of the viewport, which is to say they faded out while
|
|
1924
2449
|
still fully on screen. Journeys now run `entry calc(50% + 10vh) → exit 100%`: they start on the same
|
|
1925
2450
|
midpoint pivot as everything else and keep the full entry → hold → exit arc, so the hold sits in the
|
|
@@ -1932,7 +2457,7 @@ exit`; lightningcss parsed the end of that shorthand as `exit 0%` rather than th
|
|
|
1932
2457
|
so one knob tunes both. If your own config has an empty `slide` base, add the same line.
|
|
1933
2458
|
|
|
1934
2459
|
**Fixed: `reveal-*` on hover was unreachable.** A `hover-reveal-left` element rests at `clip-path:
|
|
1935
|
-
inset(0 100% 0 0)`, and `clip-path` clips **hit-testing** as well as painting — so it had zero
|
|
2460
|
+
inset(0 100% 0 0)`, and `clip-path` clips **hit-testing** as well as painting — so it had zero
|
|
1936
2461
|
hittable area and could never receive the hover that would reveal it. The state variants now match
|
|
1937
2462
|
the element **or its direct parent**, mirroring what `animation.css` already did for the trigger
|
|
1938
2463
|
driver (`:is(.is-active, [data-active]) > .animate-trigger`):
|