Da Content logo

Da Content

Organization
adobe
da-content

Use this when generating HTML for Adobe Document Authoring (DA, da.live) upload, uploading media binaries to DA, publishing to aem.live, or driving the DA admin API (auth, source PUT, preview/publish). Covers block HTML format (canonical div-class form and accepted table alternate), section structure, page and section metadata, icons, links, images, default content, document skeleton constraints, block cell content normalization, the DA Source API contract, IMS auth, media storage, supported formats, Media Bus vs Content Bus delivery, and silent-failure rules that corrupt content.

Overview

Publisheradobe
Repositoryskills
Skill nameda-content
Stars
298
Forks
92
Bundled files
5
LicenseApache-2.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 5 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by adobe on GitHub. Read the source before you install it.

Installation

Install the Da Content AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/adobe/skills/tree/main/plugins/aem/edge-delivery-services/skills/da-content
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/adobe/skills.git /tmp/skills
mkdir -p .claude/skills
cp -r /tmp/skills/plugins/aem/edge-delivery-services/skills/da-content .claude/skills/da-content
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Da Content in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Da Content on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Da Content is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

DA + EDS content reference

A reference skill — not a workflow. Use it whenever you need to know what the rules are for generating, uploading, or delivering content through Adobe Document Authoring (DA) and Edge Delivery Services (EDS).

This skill consolidates rules from three primary surfaces. Load the reference for the task at hand:

DoingRead
Generating HTML for DA uploadreferences/html-content.md
Uploading images, video, PDFs, fontsreferences/media.md
Hitting the DA admin API, auth, preview/publishreferences/platform.md
Re-verifying asset boundary behavior empiricallyreferences/fixtures/README.md

Every factual claim in the references is tagged [verified] (read from code or observed empirically) or [assumed] (inferred from documentation without direct verification).

When to use this skill

Invoke this skill whenever you are:

  • Generating HTML that will be uploaded to DA (admin.da.live/source/...).
  • Uploading any binary (PNG, JPG, SVG, MP4, PDF, WOFF2) to DA.
  • Calling admin.da.live (Source API) or admin.hlx.page (preview / publish API) directly.
  • Reading a DA-stored HTML document and modifying it before re-upload.
  • Diagnosing why a generated page renders incorrectly on aem.page / aem.live (silent failures: about:error images, missing meta tags, blocks rendering as plain HTML without their JS or CSS).

When NOT to use this skill

  • Writing block JS or CSS for a project — use building-blocks. This skill covers the content side, not the code side.
  • Universal Editor, structured-content authoring, or AEM Cloud Service (Java / OSGi / JCR). Out of scope.

Related Skills

  • da-auth — Obtain the IMS token needed for every admin.da.live and admin.hlx.page call. Invoke before any operation in this skill that needs DA_TOKEN.
  • generate-import-html — Generates DA-compliant HTML for imported pages; this skill is the rule reference for what that HTML must look like.
  • page-import — Orchestrates full page imports into canonical EDS block format. Reads this skill for the DA-side rules.
  • snowflake — Static-to-EDS overlay conversion. Loads this skill alongside its own methodology for DA HTML and admin API rules.

Minimal upload example

The most common operation — upload an HTML document to DA. Shows the non-obvious rules in one place: multipart/form-data body, field name data, blob with text/html type, Bearer IMS token, then a separate preview call to make the page reachable at aem.page. See references/platform.md for the full contract and references/html-content.md for what the HTML payload must look like.

bash
# Use $DA_TOKEN from the da-auth skill, or read a cached token file directly.
# See platform.md §3 for the cache locations the various Adobe DA tools use.
TOKEN="${DA_TOKEN:?invoke the da-auth skill to obtain a DA admin token}"

# 1. Upload HTML — note multipart with field name "data" (other names silently fail)
curl -X PUT \
  -H "Authorization: Bearer $TOKEN" \
  -F "data=@./page.html;type=text/html" \
  "https://admin.da.live/source/{org}/{repo}/path/to/page.html"

# 2. Trigger preview — required separate step, path WITHOUT .html extension
curl -X POST \
  -H "Authorization: Bearer $TOKEN" \
  "https://admin.hlx.page/preview/{org}/{repo}/main/path/to/page"

# 3. Optional: publish to aem.live
curl -X POST \
  -H "Authorization: Bearer $TOKEN" \
  "https://admin.hlx.page/live/{org}/{repo}/main/path/to/page"

Image binaries upload the same way (PUT to admin.da.live/source/...) but do NOT need their own preview/publish call — they're served directly from content.da.live once uploaded. They DO get pulled into Media Bus (content-addressed, with responsive variants) when a document that references them is previewed — see references/media.md §2 for the asset lifecycle.

The 11 silent-failure rules

These rules, if violated, produce broken content without any error from DA, the pipeline, or the renderer. Memorize them; verify them in generated output before upload.

  1. DA HTML is a body fragment. No <!DOCTYPE>, no <html>, no <head>, no <script>, no <style>, no inline style= attributes. The pipeline injects head/scripts/styles from Code Bus at delivery. → html-content.md §1

  2. Block class encodes block identity (canonical div form). The outermost <div> carries class="<block-name> [<variant>…]". The first class token is the block name and resolves to /blocks/<name>/<name>.{js,css}. For the accepted table-form alternate, the header is <tr><td colspan="N">Name</td></tr> where N matches the cell count of the widest content row; single-column blocks may omit colspan. Misshapen blocks (missing div class, multi-column table missing colspan, empty header cell) render as plain HTML without block JS or CSS. → html-content.md §3

  3. Block names use alphanumeric + single hyphens only. No underscores, no double dashes, no digit-first names. Variants in div form: extra class tokens after the name (class="hero cta center"). Variants in table form: parentheses after the name (Hero (cta, center)). Both normalize identically via toBlockCSSClassNames. → html-content.md §3

  4. Page Metadata block name is exactly metadata. Div form: class="metadata" (single lowercase token). Table form: header text Metadata (case-insensitive). Misspellings on either side are silently ignored — no <meta> tags emitted. → html-content.md §5

  5. <img src> URLs must be reachable from EDS preview infrastructure. The preview step fetches every <img src> and <source srcset> URL, content-hashes the bytes, and stores them in Media Bus — that's how the delivered page gets responsive <picture> variants. Any URL that doesn't return image bytes (DNS failure, 4xx/5xx, HTML response, timeout > 5s) produces <img src="about:error">. Host-less paths (repo-relative /path/foo.png, document-relative ./foo.png) also fail because the ingester has nothing to fetch. External URLs work fine and are sideloaded on first preview. → media.md §2, html-content.md §9

  6. Pre-upload binaries only when you need URL stability. Sideloading means you do NOT need to upload an image to DA before referencing it from your HTML — any reachable URL works. Pre-upload (to /media/<scope>/<file>) when you want the binary under DA's control: immune to third-party host changes, addressable by a stable content.da.live URL, and re-fetched into Media Bus on each preview. → media.md §2.5, media.md §13.2

  7. DA Source API requires multipart/form-data with field name data. Other field names (file, image) return 200 OK with no file written. → platform.md §2

  8. SVG hard cap is 40 KB. PNG/JPG/AVIF/WEBP cap is 20 MB. MP4 cap is 36 MB. Over-cap SVGs cause the preview POST to fail with 409 AEM_BACKEND_FETCH_FAILED ("Images N have failed validation"); pre-check sizes before upload. → media.md §6.1

  9. Preview / publish is a required separate step. Uploading to DA does NOT make the document visible at aem.page / aem.live. POST to admin.hlx.page/preview/... then /live/... after upload. → platform.md §6

  10. IMS tokens expire silently with 401 + empty body. Dev tokens last 24 hours. Always pre-flight expiry against expires_at in the cached token file (~/.aem/da-token.json for da-auth, .hlx/.da-token.json for @adobe/aem-cli) before a long upload run. → platform.md §3

  11. Block cell content uses stricter inline-tag normalization than default content. Inside block cells, the pipeline rewrites <b>/<i>/<s>/<mark>/<kbd> to their semantic equivalents (<strong>/<em>/<del>/<em>/<code>), strips <span> and <ins>, and applies positional rules to <br>. Visual styling survives the rewrites but CSS selectors targeting the original tags or stripped classes stop matching. Generate cell content using only the §3.9 preserve list to avoid silent reshaping. → html-content.md §3.9

Glossary

Terms used across all three references.

  • Admin API — https://admin.hlx.page/<action>/... endpoint family. Controls document lifecycle (preview, publish, status). Distinct from the DA Source API.
  • Code Bus — files delivered from the git-tracked GitHub branch (typically /fonts/, /icons/, /blocks/, /scripts/, /styles/, /head.html). Updated by code deploy.
  • Content Bus — files delivered from DA at their original path (SVG, PDF, HTML, JSON, ICO, WOFF2). Updated by preview/publish.
  • DA editor — the web UI at https://da.live/edit#/... for human authoring of documents.
  • DA Source API — https://admin.da.live/source/... endpoint for read/write of DA-tracked files (HTML and binaries).
  • Default content — anything in an EDS page outside a block: headings, paragraphs, lists, links, images. Renders as standard HTML.
  • Dot-folder — /<parent>/.<docname>/ folder created automatically by the DA editor for per-document author uploads of images.
  • EDS — Edge Delivery Services. The rendering pipeline that serves aem.page (preview) and aem.live (production), consuming DA content
    • Code Bus + Media Bus.
  • IMS token — Adobe Identity Management access token. Cached at ~/.aem/da-token.json (via the da-auth skill) or .hlx/.da-token.json (via the @adobe/aem-cli tool). Used for auth against the DA Source API and the Admin API.
  • Media Bus — content-addressed backend for image and video binaries (PNG, JPG, AVIF, WEBP, MP4). Dedup by SHA hash; permanent cache.
  • /media folder — top-level DA folder convention for shared binaries referenced across documents/branches/iterations. Auto-creates on first PUT.
  • Preview / Publish — Admin API operations that promote a document from "stored in DA" to "available at aem.page" (preview) or "available at aem.live" (publish).
  • Section — a <div> directly inside <main> in a DA HTML document. Becomes <div class="section"> after decoration.
  • Section Metadata — block whose Style key adds CSS classes to the enclosing section. NOT for SEO metadata (that's Page Metadata).
  • Page Metadata — block whose key/value rows become <head> <meta> tags at delivery. One per page; conventionally last in document.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Da Content AI skill do?

Use this when generating HTML for Adobe Document Authoring (DA, da.live) upload, uploading media binaries to DA, publishing to aem.live, or driving the DA admin API (auth, source PUT, preview/publish). Covers block HTML format (canonical div-class form and accepted table alternate), section structure, page and section metadata, icons, links, images, default content, document skeleton constraints, block cell content normalization, the DA Source API contract, IMS auth, media storage, supported formats, Media Bus vs Content Bus delivery, and silent-failure rules that corrupt content.

Why use Da Content on TypingMind?

Because you install it once and use it with any model. Da Content is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Da Content in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/adobe/skills/tree/main/plugins/aem/edge-delivery-services/skills/da-content. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Da Content?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Da Content?

As many as you like. As long as a model supports skills, you can use Da Content with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Da Content AI skill free?

Yes. It is published on GitHub by adobe under the Apache-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇