Workflow Model Design logo

Workflow Model Design

Organization
adobe
workflow-model-design

Design and create AEM Workflow models on AEM 6.5 LTS. Use when creating workflow models via the Workflow Model Editor or content package XML, defining step types (PROCESS, PARTICIPANT, DYNAMIC_PARTICIPANT, OR_SPLIT, AND_SPLIT), configuring step properties, declaring workflow variables, deploying to /etc or /conf, and syncing to /var/workflow/models via Package Manager or Maven.

Overview

Publisheradobe
Repositoryskills
Skill nameworkflow-model-design
Stars
298
Forks
92
Bundled files
8
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.

  • 8 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 Workflow Model Design 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/6.5-lts/skills/aem-workflow/workflow-model-design
  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/6.5-lts/skills/aem-workflow/workflow-model-design .claude/skills/workflow-model-design
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Workflow Model Design 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 Workflow Model Design 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 Workflow Model Design 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.

Workflow Model Design (AEM 6.5 LTS)

Design workflow models for AEM 6.5 LTS: step structure, transitions, OR/AND splits, variables, and model XML deployment via Package Manager.

Audience

Developers and workflow designers building or modifying AEM 6.5 LTS workflow models — via the Workflow Model Editor or by hand-authoring model XML and shipping it as a content package.

Variant Scope

  • This skill is AEM 6.5 LTS only.
  • Models can be stored at /conf/global/settings/workflow/models/ (preferred) or /etc/workflow/models/ (legacy — auto-deployed without Sync).
  • Not for AEM as a Cloud Service. If the target instance is AEMaaCS, stop and use the cloud-service variant of this skill — /etc/workflow/models/ is deprecated, the Sync flow differs, and several patterns documented here will produce models that fail to deploy.

Dependencies

Models reference deployed WorkflowProcess and ParticipantStepChooser services by process.label / chooser.label or fully-qualified class name. Build and deploy the Java step first; reference it from the model second. A model that references a label whose service is not yet registered fails at runtime with Process not found — Sync will succeed and the model will appear in the editor, but instances will fail on first execution. See workflow-development for step implementation.

Prerequisites

  • AEM 6.5 LTS author instance reachable (local, dev, or sandbox).
  • Maven project with ui.content module — or the ability to author in the Workflow Model Editor and export as a content package.
  • Required WorkflowProcess / ParticipantStepChooser services already implemented and deployed (see Dependencies above).
  • filter.xml covering the model path you intend to install, with mode="merge".

Required Permissions

  • workflow-editors (or equivalent write access to /conf/global/settings/workflow/models/) — create or modify models in the Workflow Model Editor or via content package.
  • workflow-administrators (or equivalent) — start test workflow instances and terminate stuck test instances during iteration.
  • Read access to /conf/global/settings/workflow/models/ and /var/workflow/models/ for verification.

Workflow

text
Model Design Progress
- [ ] 1) Clarify the workflow purpose: what triggers it, what steps are needed, who approves
- [ ] 2) Map out steps: PROCESS (auto), PARTICIPANT (human), OR_SPLIT (decision), AND_SPLIT (parallel)
- [ ] 3) Decide payload type: single JCR_PATH or multi-page workflow package (a `cq:Page` collection, detected at runtime via `adaptTo(ResourceCollection.class)`)
- [ ] 4) Identify workflow variables needed for inter-step data passing
- [ ] 5) Design model XML: flow/parsys layer with correct nt:unstructured step components and sling:resourceType per step
- [ ] 6) Choose storage: `/conf/global/settings/workflow/models/<id>` (default; requires Sync) unless the user explicitly names a site — then `/conf/<site>/settings/workflow/models/<id>`; or `/etc/workflow/models/<id>` (legacy, auto-deployed)
- [ ] 7) Add filter.xml entry with mode="merge"
- [ ] 8) Deploy via mvn install or Package Manager
- [ ] 9) Open the model in Tools → Workflow → Models → Edit → Sync; verify the model appears in /var/workflow/models/ and all steps render on the editor canvas
- [ ] 10) Start a test workflow instance; confirm it runs to completion

Output Contract

Generate only these files for a /conf-based model:

FileNode typePurpose
jcr_root/conf/.../models/<id>/.content.xmlcq:PageModel root page
jcr_root/conf/.../models/<id>/jcr:content/.content.xmlcq:PageContentTitle, template, resourceType
jcr_root/conf/.../models/<id>/jcr:content/flow/.content.xmlnt:unstructured + parsysStep nodes

Never generate — hard stops:

  • ❌ Any file under jcr_root/var/ — AEM Sync writes /var/workflow/models/ automatically after you click Sync in the editor. Never ship /var content in a content package.
  • ❌ A model/ directory inside jcr:content/ at the /conf path — a cq:WorkflowModel node with <nodes> and <transitions> is the runtime format. It must never appear under /conf. The Workflow Model Editor cannot open a model that contains it.
  • ❌ <filter root="/var/workflow/models/..."/> in filter.xml — /var is never a package filter target. The only filter entry needed is the design-time /conf (or /etc) path.

filter.xml — correct entry:

xml
<filter root="/conf/global/settings/workflow/models/my-workflow" mode="merge"/>

Node Types Quick Reference

TypePurposeKey metaData property
STARTEntry point—
ENDTerminal—
PROCESSAuto-executed Java stepPROCESS = FQCN or process.label
PARTICIPANTHuman task (static assignee)PARTICIPANT = principal name
DYNAMIC_PARTICIPANTHuman task (runtime assignee)DYNAMIC_PARTICIPANT = chooser.label
OR_SPLITOne branch selected by ruleTransition rule = ECMAScript (Rhino)
AND_SPLITAll branches execute in parallel—
AND_JOINWait for all parallel branches—

Default Path Rule

Unless the user explicitly names a specific AEM site, always generate models at /conf/global/settings/workflow/models/<id>. Do not infer a site-scoped path (e.g., /conf/wknd/…) from conversational context such as "for the WKND site" or "install on the WKND instance." Workflow models are not site-scoped by default — they are global resources. Only use /conf/<site>/settings/workflow/models/<id> when the user explicitly states that the model should be scoped to a specific site.

6.5 LTS: /conf vs /etc

LocationSync RequiredBest For
/conf/global/settings/workflow/models/Yes — via UI or deployModel()New models, forward-compatible
/etc/workflow/models/No — auto-deployedLegacy models, simple setups

Design-time vs runtime paths: /conf/global/settings/workflow/models/ stores the editor's flow/parsys format — this is what you author and deploy in a content package. Steps are nt:unstructured nodes with sling:resourceType; no cq:WorkflowNode or cq:WorkflowTransition nodes appear here. /var/workflow/models/ holds the runtime copy generated by Sync — it uses cq:WorkflowModel, cq:WorkflowNode, and cq:WorkflowTransition. Never write cq:WorkflowModel nodes to the /conf path, and never hand-author /var content — AEM manages the runtime layer entirely via Sync.

Deployment

bash
# Deploy via Maven
mvn clean install -P autoInstallPackage

# Or install package manually via Package Manager
http://localhost:4502/crx/packmgr

Architecture Considerations

These choices are made at design time and are hard to retrofit. Apply them before locking the model shape:

  • Transient vs persistent. Set transient="true" on the model for high-volume, short-lived workflows (e.g., asset metadata extraction, replication side-effects). Transient instances create no JCR node under /var/workflow/instances until a retry or external-process step forces persistence — the difference between a healthy repository and unbounded growth.
  • Participant timeouts. A PARTICIPANT step with no timeout can stall a workflow indefinitely and consume a job slot for the duration. For any human task that may not be acted on promptly, configure TIMEOUT and a TIMEOUT_HANDLER route on the step's metaData.
  • Goto retry caps. Always enforce a hard cap on retry counters used by Goto Step rules (e.g., count < 3). An uncapped Goto loop will pin a worker thread and accumulate failed instances until the workflow is terminated by hand.
  • Purge is a design-time concern. Estimate instance volume before go-live. Configure the Adobe Granite Workflow Purge Configuration factory in the same content package as the model so purge ships with the workflow, not as an after-the-fact follow-up. Set scheduledpurge.purgePackagePayload=true if the workflow uses workflow packages.
  • Model versioning. In-flight instances stay bound to the model version they started with; new instances pick up the new version. When changing a deployed model, plan a brief drain window for old-version instances before declaring the change complete.

References

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 Workflow Model Design AI skill do?

Design and create AEM Workflow models on AEM 6.5 LTS. Use when creating workflow models via the Workflow Model Editor or content package XML, defining step types (PROCESS, PARTICIPANT, DYNAMIC_PARTICIPANT, OR_SPLIT, AND_SPLIT), configuring step properties, declaring workflow variables, deploying to /etc or /conf, and syncing to /var/workflow/models via Package Manager or Maven.

Why use Workflow Model Design on TypingMind?

Because you install it once and use it with any model. Workflow Model Design 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 Workflow Model Design in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/adobe/skills/tree/main/plugins/aem/6.5-lts/skills/aem-workflow/workflow-model-design. 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 Workflow Model Design?

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 Workflow Model Design?

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

Is the Workflow Model Design 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 👇