docs: update numbers

AGENTS.md

@@ -14,7 +14,7 @@ This file provides guidance to AI coding agents when working with code in this r
 - `pnpm lint` — ESLint (flat config in `eslint.config.mjs`, includes `eslint-plugin-jsdoc`).
 - `pnpm format` — Prettier over the repo.
 - `pnpm build` — `tsc` emits `.d.ts` types, then `copy:source` uses esbuild to minify/bundle `src/*.js` + `src/utils/*` into `dist/` as ESM. **`dist/` is what gets published; `src/` is shipped as the readable source.**
-- `pnpm size-limit` — enforces the per-file byte budgets declared in `package.json` `size-limit` (each `dist/` file has its own, e.g. `WebComponent.js` ≤ 2.05 KB, `html.js` ≤ 0.6 KB, each util ≤ 0.5–0.8 KB). Keep additions tiny; this gate is a core project value. `size-change-log.md` is the running record of how each change moved the base-class bundle and why.
+- `pnpm size-limit` — enforces the per-file byte budgets declared in `package.json` `size-limit` (each `dist/` file has its own, e.g. `WebComponent.js` ≤ 2.05 KB, `html.js` ≤ 0.6 KB, each util ≤ 0.5–0.8 KB). Keep additions tiny; this gate is a core project value. `size-change-log.md` is the running record of how each change moved the base-class bundle and why; its latest row's min+brotli figure is the single source of truth for the base-class size, so `docs/src/content/docs/guides/library-size.md` must be updated in lockstep whenever a new row lands.
 - `pnpm test:types` — builds, then type-checks the TS type tests (`test/types/typed-props.test-d.ts`) and the `demo/examples/typed-props/` example. `this.props` is typed by passing the props shape as a class type argument: `class X extends WebComponent<typeof props>`.
 - `pnpm docs` — run the Astro docs site (`docs/` workspace) locally.
 - `pnpm demo` — run the Vite examples showcase (`demo/` workspace) locally; `pnpm demo:build` builds it.
@@ -54,7 +54,7 @@ Any change to observable behavior ships as one unit — code alone is an incompl
 1. **Tests** — unit specs in `test/` (or colocated `*.test.mjs`) covering the new contract *and* the old behavior it replaces. Add a `test/e2e/` spec whenever the behavior depends on something happy-dom cannot model faithfully — CSS selector matching, computed styles, custom-element upgrade timing.
 2. **Demo examples** — a runnable example under `demo/examples/` that demonstrates the behavior, linked from a card in `demo/index.html`. Update any existing example the change affects, including ones that now emit a console warning or model a discouraged pattern.
 3. **Documentation** — the guide under `docs/src/content/docs/guides/` that doubles as the behavioral spec. For a breaking change, also update the `README.md` banner with the migration consumers have to perform.
-4. **Size budget** — `pnpm size-limit` stays green. If an addition genuinely needs more headroom, raise the budget in `package.json` deliberately and say so in the change description; never let it drift silently.
+4. **Size budget** — `pnpm size-limit` stays green. If an addition genuinely needs more headroom, raise the budget in `package.json` deliberately and say so in the change description; never let it drift silently. Whenever the change moves the base-class bundle, record it as a new row in `size-change-log.md` **and** update the headline size in `docs/src/content/docs/guides/library-size.md` to match that row's new min+brotli figure — the two must never disagree.
 
 Verify with `pnpm test:all` (unit + types + e2e across all engines) before calling the change done.
 

docs/src/components/SizeChart.astro

@@ -6,7 +6,7 @@
 const data = [
   { name: 'vanilla', kb: 0.2 },
   { name: 'web-component-base', kb: 2.6, accent: true },
-  { name: '@elenajs/core', kb: 3.3 },
+  { name: '@elenajs/core', kb: 3.4 },
   { name: 'lit', kb: 5.3 },
   { name: 'fast-element', kb: 12.2 },
 ]
@@ -15,7 +15,7 @@ const max = Math.max(...data.map((d) => d.kb))
 
 <figure
   class="viz-root"
-  aria-label="Bundle size of the same counter component per library, in kilobytes, minified and brotli-compressed. Vanilla 0.2, web-component-base 2.6, elenajs core 3.3, lit 5.3, fast-element 12.2."
+  aria-label="Bundle size of the same counter component per library, in kilobytes, minified and brotli-compressed. Vanilla 0.2, web-component-base 2.6, elenajs core 3.4, lit 5.3, fast-element 12.2."
 >
   <div class="chart">
     {

docs/src/content/docs/guides/comparison.md

@@ -15,10 +15,10 @@ See it live: [Library comparison demo ↗](https://demo.webcomponent.io/examples
 
 | Library                   | Version | Minified | Gzip    | Brotli     |
 | ------------------------- | ------- | -------- | ------- | ---------- |
-| **web-component-base**    | 6.1.0   | 6.7 kB   | 2.9 kB  | **2.6 kB** |
-| `@elenajs/core`           | 1.0.0   | 9.1 kB   | 3.7 kB  | 3.3 kB     |
+| **web-component-base**    | 6.1.1   | 6.7 kB   | 2.9 kB  | **2.6 kB** |
+| `@elenajs/core`           | 1.0.0   | 9.1 kB   | 3.7 kB  | 3.4 kB     |
 | `lit`                     | 3.3.3   | 15.3 kB  | 5.9 kB  | 5.3 kB     |
-| `@microsoft/fast-element` | 3.0.1   | 44.9 kB  | 13.7 kB | 12.2 kB    |
+| `@microsoft/fast-element` | 3.0.1   | 44.8 kB  | 13.6 kB | 12.2 kB    |
 | vanilla `HTMLElement`     | -       | -        | -       | ~0.2 kB    |
 
 For scale: even after all of the v5.2–v6.1 work, the WCB counter is **~21% smaller than Elena, ~51% smaller than Lit, and ~79% smaller than FAST**.
@@ -51,4 +51,4 @@ For what these numbers and capabilities add up to (and when they don't) see [Why
 
 ---
 
-_WCB re-measured 2026-07-20 at v6.1.0; the other libraries measured 2026-07-19, with esbuild, Node zlib (gzip −9, brotli q11), at the pinned versions above. Methodology: identical counter component per library, bundled per library, compressed. Re-run them yourself — the counters and the [`measure.mjs`](https://demo.webcomponent.io/examples/library-comparison/) script live in the demo workspace (`demo/examples/library-comparison/`). The benchmark is trivially reproducible with the versions pinned above._
+_WCB re-measured 2026-07-23 at v6.1.1; the other libraries measured 2026-07-19, with esbuild, Node zlib (gzip −9, brotli q11), at the pinned versions above. Methodology: identical counter component per library, bundled per library, compressed. Re-run them yourself — the counters and the [`measure.mjs`](https://demo.webcomponent.io/examples/library-comparison/) script live in the demo workspace (`demo/examples/library-comparison/`). The benchmark is trivially reproducible with the versions pinned above._

docs/src/content/docs/guides/why.md

@@ -11,7 +11,7 @@ WCB exists for five reasons:
 2. **Zero tooling, genuinely.** No compiler, no decorators, no build step: one `import` from a CDN in a `<script type="module">` works on current browsers. The whole mental model is `static props` + `template` + four hooks, and the shipped source is readable in one sitting. The runtime is 2.6 kB brotli-compressed ([measured](/comparison/)).
 3. **Attribute-first reactivity is HTML-native.** Because props serialize to attributes, initial state can be rendered by _any_ server in plain HTML (no JS-framework SSR integration needed) and components stay inspectable/debuggable in devtools as ordinary attributes.
 4. **Light DOM by default.** Global stylesheets, forms, and third-party CSS just work; and the shadow DOM is one static field away when you want encapsulation.
-5. **The size gate is a governed value.** Every byte added must justify itself in the [size change log](https://github.com/ayo-run/wcb/blob/main/size-change-log.md), enforced by `size-limit` budgets in CI. Smart diffing is the largest single addition in the project's history and it cost 0.42 kB.
+5. **The size gate is a governed value.** Every byte added must justify itself in the [size change log](https://github.com/ayo-run/wcb/blob/main/size-change-log.md), enforced by `size-limit` budgets in CI. Smart diffing is the largest single addition in the project's history and it cost 0.43 kB.
 
 ## When WCB is the wrong choice