Skills for building AEM Edge Delivery Services sites — site creation, block development, content modeling, code review, testing, and page import.
91%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Low
Low-risk findings worth noting
Turn a Figma design into an Adobe Edge Delivery Services
(EDS / AEM / Franklin / Helix) content page and deploy it to Document
Authoring (da.live). Give it a design and the skill infers each section's
mapping against the project's existing blocks and confirms the plan with you
before building — no annotations needed ("just migrate this page"). If a design
happens to be annotated, those annotations are used as an optional
accelerator, but they are never required.
Runs standalone in plain Claude / Coworker with just a Figma MCP (to read the design) and a DA IMS token (to write content) — no proprietary web app or migration tooling required.
The skill classifies each section of the design and follows the matching path:
A single design usually mixes both, plus default content.
preflight: Figma MCP reachable · file access · DA token · project checkout (fail fast)
▼
Figma frame (annotations optional)
│ Figma MCP (get_metadata, get_design_context, get_screenshot, get_variable_defs)
▼
section inventory ──► resolve each section (infer + confirm; annotations optional)
│ ├─ existing block (structure + visual fit) ─► author content
│ ├─ needs one / look diverges → new block ───► build isolated block
│ │ (content-driven-development), push code
│ └─ default content ──────────────────────────► author prose/media, no block
▼
resolution plan ──► confirm with user (ask on low-confidence; flag overwrite of an existing page)
▼
content/<path>.html ──► PUT admin.da.live/source ──► POST admin.hlx.page/preview
(da-content contract) (+ Code Sync first, for new blocks)
▼
pre-publish gate: legibility · no placeholder · icons resolve · blocks decorate · metadata present
▼
da.live edit URL + aem.page preview URLThe flow is bracketed by two safety gates: a preflight (fail fast on a
missing Figma MCP, no file access, no DA token, or no project checkout — one
actionable message instead of a half-built page) and a pre-publish gate
(the page is not "done" until legibility, placeholder-free content, resolvable
icons, decorated new blocks, and a populated metadata block are all verified —
an un-run check counts as failed, never as a pass).
It orchestrates existing public skills rather than reinventing them — the net-new logic is reading the Figma design, resolving each section to a block (inferred by structure + visual fit and confirmed with you, or from annotations when the frame has them), and mapping content into it. DA rules, auth, and block-building all defer to the skills below.
skills/figma-to-content/
├── SKILL.md Entry point (agent reads first)
├── README.md This file
├── package.json
└── references/
└── annotation-contract.md The annotation spec (optional path; proposed)da-auth — DA IMS tokenda-content — DA HTML rules + Source API + preview/publish + mediablock-collection-and-party — whether a block exists + its authoring modelcontent-modeling — design a content model for a new blockcontent-driven-development — full dev workflow for a reusable new block (invokes building-blocks, testing-blocks)testing-blocks — render a block + screenshot and compare to the design; reused directly for the reuse gate's visual checkDraft (v0.1.0). The infer-and-confirm path is the primary, fully working
flow. Before v1.0.0, close the open questions in SKILL.md (chiefly the default
image-hosting choice). The annotation contract
(references/annotation-contract.md) is
an optional enhancement for teams that want to pre-declare section mappings
— nice to have, not required.
Intended destination:
adobe/skills → plugins/aem/edge-delivery-services/skills/figma-to-content.
.tessl-plugin
skills
aem-block-library
aem-cli
aem-layout-view
block-collection-and-party
brand-setup
references
building-blocks
code-review
content-driven-development
references
content-modeling
create-site
da-auth
figma-to-content
find-test-content
generate-import-html
identify-page-structure
migrate-header
page-decomposition
page-import
preview-import
scrape-webpage
slicc-handoff
snowflake
assets
substrate
substrate-milo
examples
knowledge
scripts
testing-blocks
ue-component-model