AGENT SKILL

pixijs-html-source

Use this skill when rendering live HTML/DOM elements (or frozen snapshots of them) as PixiJS v8 textures via the EXPERIMENTAL HTML-in-Canvas browser APIs. Covers the pixi.js/html-source import, feature detection with canvas.requestPaint, HTMLSource for a live, repainting, still-interactive element (autoLayout/autoUpdate/autoRequestPaint, requestPaint, isReady, HTMLSource.defaultOptions, the direct-child-of-canvas + layoutsubtree requirement), ElementImageSource for an immutable captureElementImage() snapshot (autoClose, ready immediately), using the source on a Sprite/Texture/Mesh, fallback-only auto-detection via Texture.from at priority -10, uploader error messages, and destroy/cleanup. Triggers on: HTMLSource, ElementImageSource, ElementImage, HTML in canvas, render DOM to texture, render HTML to texture, texElementImage2D, copyElementImageToTexture, HTMLSourceOptions, ElementImageSourceOptions, HTMLSourceCanvas.

License 许可
MIT
所属仓库
bg-szy/TOP-SKILLS
最近更新
2026-11-15
Stars
★ 6
Forks
0

01 前言 Front Matter

name
pixijs-html-source
license
MIT
description
Use this skill when rendering live HTML/DOM elements (or frozen snapshots of them) as PixiJS v8 textures via the EXPERIMENTAL HTML-in-Canvas browser APIs. Covers the pixi.js/html-source import, feature detection with canvas.requestPaint, HTMLSource for a live, repainting, still-interactive element (autoLayout/autoUpdate/autoRequestPaint, requestPaint, isReady, HTMLSource.defaultOptions, the direct-child-of-canvas + layoutsubtree requirement), ElementImageSource for an immutable captureElementImage() snapshot (autoClose, ready immediately), using the source on a Sprite/Texture/Mesh, fallback-only auto-detection via Texture.from at priority -10, uploader error messages, and destroy/cleanup. Triggers on: HTMLSource, ElementImageSource, ElementImage, HTML in canvas, render DOM to texture, render HTML to texture, texElementImage2D, copyElementImageToTexture, HTMLSourceOptions, ElementImageSourceOptions, HTMLSourceCanvas.

02 资源文件 1

03 SKILL.md 正文

HTMLSource and ElementImageSource turn a DOM element into a TextureSource you can use anywhere a normal texture works: on a Sprite, as a Texture frame, or mapped onto a Mesh. HTMLSource mirrors a live element's pixels into the GPU (the element stays editable and clickable in the browser); ElementImageSource wraps an immutable snapshot that never repaints. Both live in pixi.js/html-source; importing from that path registers their extensions.

These sources rely on the experimental HTML-in-Canvas browser proposal and are marked EXPERIMENTAL in PixiJS v8. The browser API must be enabled or the texture uploader throws on first render; feature-detect with canvas.requestPaint before relying on it. The API may change between minor releases.

Assumes familiarity with pixijs-scene-sprite and textures. These are texture sources, not display objects: wrap them in a Sprite (or Texture/Mesh) to put them on screen. Not available in Web Workers; a worker has no DOM to capture.

Quick Start

import { Application, Sprite } from "pixi.js";
import { HTMLSource } from "pixi.js/html-source";

const app = new Application();
await app.init({ resizeTo: window });
document.body.appendChild(app.canvas);

// The element must be a direct child of the Pixi canvas.
const form = document.createElement("form");
form.innerHTML = '<input value="still editable" />';
app.canvas.appendChild(form);

// Render the live form as a sprite. It stays interactive in the browser.
const source = new HTMLSource({ resource: form });
const sprite = Sprite.from(source);

sprite.anchor.set(0.5);
sprite.position.set(app.screen.width / 2, app.screen.height / 2);
app.stage.addChild(sprite);

Related skills: pixijs-scene-sprite (display the texture), pixijs-scene-mesh (map onto geometry, PerspectiveMesh), pixijs-scene-dom-container (the opposite: overlay HTML above the canvas, outside the GPU pipeline), pixijs-assets (texture sources vs the loader/cache), pixijs-environments (no DOM in Web Workers).

Constructor options

Both sources extend TextureSource, so all TextureSourceOptions (resolution, scaleMode, addressMode, label, etc.) are valid. resource is required on each.

HTMLSourceOptions (live element):

Option Type Default Description
resource Element — Required. The live DOM element to render. Must be a direct child of the owning canvas, or the constructor throws.
canvas HTMLSourceCanvas — The canvas that owns the element's layout subtree. Inferred from resource.parentElement when the element is a direct canvas child; pass it when inference is not possible.
autoLayout boolean true Set the layoutsubtree attribute on the owning canvas. The browser only lays out and paints canvas children when it is present. Set false if you write <canvas layoutsubtree> yourself.
autoUpdate boolean true Listen for the canvas paint event and re-upload when the element repaints. Set false for a static, captured-once texture.
autoRequestPaint boolean true Request one initial paint after construction. Set false and call source.requestPaint() yourself each frame for continuous animation.

ElementImageSourceOptions (immutable snapshot):

Option Type Default Description
resource ElementImage — Required. A snapshot from canvas.captureElementImage(element).
autoClose boolean false Call snapshot.close() when the source is destroyed. Leave false when the snapshot is shared with other sources, or you risk a use-after-free.

Change the defaults for every HTMLSource through the static HTMLSource.defaultOptions object, for example HTMLSource.defaultOptions.autoUpdate = false.

Core Patterns

Setup and the side-effect import

import "pixi.js/html-source";
import { HTMLSource, ElementImageSource } from "pixi.js/html-source";

pixi.js/html-source calls extensions.add(...) to register HTMLSource, ElementImageSource, and their WebGL/WebGPU uploaders. Without it, the renderer has no 'html' uploader and these sources never render. The classes are exported from pixi.js/html-source, not pixi.js.

Importing a named export from pixi.js/html-source also triggers the side effect, so a bare import 'pixi.js/html-source' is only needed when you don't import anything else from that path.

Feature detection and browser support

The HTML-in-Canvas API is gated behind a browser flag. Feature-detect before relying on it:

import type { HTMLSourceCanvas } from "pixi.js/html-source";

const canvas = app.canvas as HTMLSourceCanvas;

if (canvas.requestPaint) {
  // HTML-in-Canvas is available.
}

Cast app.canvas to HTMLSourceCanvas for the typed requestPaint and captureElementImage members. source.requestPaint() returns false when the browser lacks the API; the texture uploader throws on first render when it is disabled.

When the API is missing the first upload throws [HTMLSource] WebGLRenderingContext.texElementImage2D is not available... on WebGL or [HTMLSource] GPUQueue.copyElementImageToTexture is not available... on WebGPU.

Live element with HTMLSource

const form = document.createElement("form");
app.canvas.appendChild(form); // direct child of the canvas

const source = new HTMLSource({ resource: form });
const sprite = Sprite.from(source);

The element must be a direct child of the renderer's <canvas>; the source infers the owning canvas from resource.parentElement (or pass canvas). With the defaults, it sets layoutsubtree on the canvas, listens for the canvas paint event, and requests one initial paint. source.isReady is false until that first paint lands, then true. It is true immediately when autoUpdate is false. resourceWidth/resourceHeight report the element's real-pixel size (offsetWidth/offsetHeight).

Continuous animation with requestPaint

const source = new HTMLSource({ resource: clock, autoRequestPaint: false });
const sprite = Sprite.from(source);

app.ticker.add(() => {
  clock.textContent = new Date().toLocaleTimeString();
  source.requestPaint(); // re-snapshot the DOM this frame
});

The browser only repaints canvas children on demand. For an element whose content changes every frame, set autoRequestPaint: false and call source.requestPaint() in your own ticker to drive repaints on your schedule.

Immutable snapshot with ElementImageSource

import { ElementImageSource } from "pixi.js/html-source";
import type { HTMLSourceCanvas } from "pixi.js/html-source";

const canvas = app.canvas as HTMLSourceCanvas;
const snapshot = canvas.captureElementImage!(element);

const source = new ElementImageSource({ resource: snapshot, autoClose: true });
const sprite = Sprite.from(source);

captureElementImage() freezes an element's current pixels into an immutable ElementImage. There is no owning canvas, no paint listener, and no repaint lifecycle, so the source is ready the moment it is constructed. Reach for it when you need a frozen copy that outlives its element or is transferred around (transitions, "shatter" or trail effects). Release the snapshot with snapshot.close() when done, or pass autoClose: true to let the source close it on destroy().

Using the source on a sprite, texture, or mesh

Both sources are normal TextureSources. Wrap them with Sprite.from(source) / Texture.from(source), frame or slice them into sub-textures, or map them onto a mesh:

import { Rectangle, Texture } from "pixi.js";

// A 64x64 slice of the rendered element.
const chunk = new Texture({
  source,
  frame: new Rectangle(0, 0, 64, 64),
});

// Mapped onto geometry (e.g. a perspective warp).
const mesh = new PerspectiveMesh({ texture: Texture.from(source) /* ... */ });

Auto-detection and priority

// Resolves to an HTMLSource (element) or ElementImageSource (snapshot) only as a last resort.
const sprite = Sprite.from(elementAlreadyInTheCanvas);

A generic HTML element or an ElementImage passed to Texture.from/Sprite.from resolves to these sources at the lowest texture-source priority (-10), so they only claim a resource no other built-in source handles. Image, video, and canvas elements are deliberately rejected; they have dedicated, faster sources. Construct the source explicitly when you need options (autoUpdate, autoClose) or non-HTML elements such as SVG.

Common Mistakes

[HIGH] Importing from pixi.js instead of pixi.js/html-source

Wrong:

import { HTMLSource } from "pixi.js"; // not exported from the core entry

Correct:

import { HTMLSource } from "pixi.js/html-source";

HTMLSource, ElementImageSource, and the 'html' uploaders live only in pixi.js/html-source. Importing anything from that path registers the extensions; the module is marked side-effectful, so bundlers keep the registration even when you only import a class.

[HIGH] Assuming the browser API is enabled

The HTML-in-Canvas proposal is not shipped by default. If the API is disabled, the uploader throws on first render. Feature-detect first:

const canvas = app.canvas as HTMLSourceCanvas;
if (!canvas.requestPaint) {
  // Fall back to a static image, DOMContainer overlay, or a message.
}

[MEDIUM] Element not a direct child of the canvas

document.body.appendChild(form); // wrong parent
const source = new HTMLSource({ resource: form }); // throws

HTMLSource requires the element to be a direct child of the owning canvas (app.canvas.appendChild(form)) and throws otherwise. Append the element to the canvas before constructing the source, or pass the canvas option.

[MEDIUM] Expecting a live element to update without requestPaint

A non-animating element updates automatically on browser paint events (autoUpdate: true). Content that changes every frame will not re-upload unless something triggers a paint; drive it with source.requestPaint() each frame (with autoRequestPaint: false).

[MEDIUM] Closing an ElementImage still in use

const source = new ElementImageSource({ resource: snapshot, autoClose: true });
const other = new ElementImageSource({ resource: snapshot }); // shares the snapshot

source.destroy(); // closes the snapshot — `other` is now a use-after-free

Only set autoClose: true when the source owns the snapshot exclusively. For a shared snapshot, leave autoClose off and call snapshot.close() once, after the last source is destroyed.

Cleanup

source.destroy();

HTMLSource.destroy() detaches the canvas paint listener and nulls its canvas reference. ElementImageSource.destroy() closes the snapshot when autoClose was set; otherwise call snapshot.close() yourself to release the memory.

API Reference

04 同仓库更多 Skills

bg-szy
bg-szy/TOP-SKILLS

serpmaven

Scrape Google search results (SERP) to JSON with SERP Maven (serpmaven) - a real Chrome browser on a residential proxy with sticky sessions. Run it as a CLI, a Python library, a local HTTP search API (`serve`), an MCP server or a Docker image; set the proxy login, workers, pages and cost tracking. Use when the user wants Google results, rank tracking, a self-hosted SERP API, or live Google search for an AI agent.

★ 6
bg-szy
bg-szy/TOP-SKILLS

revnu-landing-page-audit

Find out why a landing page isn't turning visitors into signups, demos or sales. Audits one page against what the ad, post or search that sends people there promised, then checks the first screen, proof, call to action, friction and mobile. Returns at most 7 prioritized fixes, plus what is fine and what could not be checked; rewrites specific lines and fixes the page in the repo when asked. Never promises a conversion lift. Works without a Revnu account. Use when the user says or asks: "audit my landing page", "why isn't this page converting", "review my homepage", "my ads get clicks but no signups", "people bounce from the pricing page", "CRO review", "improve conversion rate", "is my hero section clear", "roast my landing page", "rewrite my headline", "what's wrong with this page", "landing page feedback", "make the signup page convert better". Not for SEO or AI-search audits, page speed engineering on its own, or building a new page from scratch.

★ 6
bg-szy
bg-szy/TOP-SKILLS

revnu-growth-context

Write down once who buys, what they're stuck on, what it costs, what proof you have and what you've already tried, so every growth task starts from facts instead of a questionnaire. Creates or updates .agents/revnu-growth.md from the repo, the live site and the user, and imports an existing product-marketing.md. Use when the user says "set up our marketing context", "who is our ideal customer", "write our ICP", "describe the product for marketing", "positioning doc", "what's our brand voice", "remember this about our customers", "log what we tried", "what have we learned from marketing", "we changed pricing, update the context", "update our growth notes", "summarize the business for the agent", "import my product-marketing.md", "start a growth doc", or before a first marketing, SEO, outreach or launch task when no context file exists. Works without a Revnu account.

★ 6
bg-szy
bg-szy/TOP-SKILLS

revnu-get-listed

Get onto the third-party pages AI answers and search cite when buyers ask for the best tool in your category: "best X" roundups, directories, review sites, GitHub awesome lists, community threads and comparison pages. Finds which ones are cited, ranks them by fit and effort, and drafts each submission, PR or pitch by the target's own rules. Use when the user says "get us listed", "which directories should we submit to", "how do we get into best-of lists", "why does ChatGPT recommend competitors and not us", "get mentioned in AI answers", "submit to awesome lists", "get on G2", "where are our competitors listed", "pitch listicle authors", "find backlink opportunities", "launch directories", "get into roundups", "Product Hunt alternatives", "where should we be listed", "get cited by Perplexity". Never submits, pays or posts without a yes. Works without a Revnu account.

★ 6
bg-szy
bg-szy/TOP-SKILLS

revnu-ai-readiness-audit

Audit whether your website gives ChatGPT, Perplexity, Claude, Gemini and Google's AI Overviews something worth quoting, and fix the pages that don't. Use when the user says or asks "is my site ready for AI search", "audit my site for AI answers", "AEO audit", "GEO audit", "why doesn't ChatGPT quote our pricing", "make our pages citable by AI", "do I need llms.txt", "does schema help with ChatGPT", "check our structured data", "is our content visible without JavaScript", "AI SEO audit", "optimize our homepage for AI answers", "LLM SEO check", "why does AI get our product wrong", "fix our site for AI Overviews". Checks fetchability and indexability, answers in the raw HTML, entity facts, schema (including JS-injected markup), headings, dates and llms.txt, then gives at most 10 prioritised fixes and edits the repo when asked. Works without a Revnu account.

★ 6
bg-szy
bg-szy/TOP-SKILLS

proxy-troubleshooting

Diagnose residential proxy failures, with NodeMaven's gateway as the reference - 407 with correct credentials, 406, 410, 500, a connection that hangs, ERR_TUNNEL_CONNECTION_FAILED in a browser, Playwright or Puppeteer ignoring proxy credentials, a sticky session that changes IP, geo-targeting that is silently ignored, passwords with special characters, and unexpected traffic from headless Chrome. Use when a proxy connection fails, returns the wrong location, or costs more traffic than expected.

★ 6