Workflow Launchers logo

Workflow Launchers

Organization
adobe
workflow-launchers

Configure and deploy Workflow Launchers that automatically start workflows in response to JCR content changes on AEM 6.5 LTS

Overview

Publisheradobe
Repositoryskills
Skill nameworkflow-launchers
Stars
298
Forks
92
Bundled files
2
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.

  • 2 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 Launchers 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-launchers
  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-launchers .claude/skills/workflow-launchers
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Workflow Launchers 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 Launchers 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 Launchers 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 Launchers Skill — AEM 6.5 LTS

Audience

Developers and integrators configuring cq:WorkflowLauncher nodes that auto-start workflows on JCR content events on AEM 6.5 LTS — DAM asset processing on upload, review workflows on page edit, custom auto-trigger patterns, or overlays that disable/replace OOTB launcher behavior.

Variant Scope

  • AEM 6.5 LTS only.
  • Custom launchers live at /conf/global/settings/workflow/launcher/config/ (preferred) or /apps/settings/workflow/launcher/config/. Legacy /etc/workflow/launcher/config/ still works but should be migrated.
  • Not for AEM as a Cloud Service. AEMaaCS uses Sling Job Topics for event distribution; the JCR-event-listener launcher pattern documented here does not apply directly. If the target is AEMaaCS, stop and use the cloud-service variant of this skill.

Dependencies

Launchers depend on three upstream concerns — verify all three before expecting a launcher to start a working instance:

  • workflow-model-design — the workflow referenced by the launcher's workflow= property must already be deployed and synced to /var/workflow/models/<name>.
  • workflow-development — every WorkflowProcess and ParticipantStepChooser referenced by that model must be registered as an OSGi service. Missing services produce Process not found on first instance execution.
  • workflow-triggering — launchers are one of several triggering mechanisms; if you need a different one (manual, programmatic, HTTP API), see workflow-triggering.

Prerequisites

  • AEM 6.5 LTS author instance reachable.
  • Workflow model deployed and visible at /var/workflow/models/<name> (verify via Tools → Workflow → Models).
  • For OOTB-launcher overlays: write access to /conf/global/ or /apps/settings/.
  • filter.xml covering the launcher path with mode="merge".

Required Permissions

  • Write access to /conf/global/settings/workflow/launcher/config/ (or /apps/settings/...) for deploying custom launchers via content package.
  • workflow-administrators (or equivalent) — enable/disable launchers in the Tools → Workflow → Launchers UI.
  • Read access to /var/workflow/models/ for runtime path lookup.

Common Scenarios

Use this table to route a developer's intent to the right launcher pattern:

Developer intentPattern
"Start a workflow when an asset is uploaded to DAM"NODE_ADDED on nt:file under /content/dam(/.*)?/jcr:content/renditions/original
"Trigger review when a page is edited under /content/my-site"NODE_MODIFIED on cq:PageContent under /content/my-site(/.*)?/jcr:content
"Disable an OOTB DAM launcher I don't need"Overlay at /conf/global/.../launcher/config/<same-name> with enabled={Boolean}false
"Replace OOTB launcher behavior with my own workflow"Overlay at /conf/global/.../launcher/config/<same-name> with new workflow=
"Auto-start on every replication event"Use a Replication Trigger instead — see workflow-triggering Section 5

When NOT to Use a Launcher

Launchers are event-based (JCR observation). Use a different mechanism when the trigger is not an event-driven content change:

Developer intentUse this instead
"Run my workflow nightly" or any time-based scheduleSling Scheduler + WorkflowSession API — see workflow-triggering programmatic-api.md
"Run this once for all 500 existing pages"A one-shot servlet or scheduled job that starts workflows in a capped loop
"When workflow X completes, run workflow Y"A WorkflowProcess step in workflow X that triggers Y, not a second launcher (avoids race conditions)
"Run when content is replicated to publish"Replication Trigger (workflow-triggering Section 5), not a launcher on /var/audit/...
"Run on events under /var/, /jcr:system, or anonymous-user events"These paths and the anonymous user are excluded by WorkflowLauncherListener — launchers cannot fire here

Core Concept: What Is a Workflow Launcher?

A Workflow Launcher (cq:WorkflowLauncher) is a JCR node that registers a JCR event listener. When a node event occurs at a path matching the launcher's glob pattern, node type, and conditions, the Granite Workflow Engine enqueues a workflow start.

The listener is managed by WorkflowLauncherListener (an OSGi service). It reads all active launcher configurations at startup and re-evaluates them when configurations change.

Architecture at a Glance

JCR Event (NODE_ADDED / NODE_MODIFIED / NODE_REMOVED)
    ↓
WorkflowLauncherListener (OSGi EventListener)
    ↓ matches: glob, nodetype, event type, conditions
Workflow Engine: enqueue WorkflowData
    ↓
Workflow Instance created at /var/workflow/instances/

Launcher Configuration Properties

PropertyTypeDescription
eventTypeLong1 = NODE_ADDED, 2 = NODE_MODIFIED, 4 = NODE_REMOVED, 8 = PROPERTY_ADDED, 16 = PROPERTY_CHANGED, 32 = PROPERTY_REMOVED
globStringGlob pattern matched against the event node path (e.g., /content/dam(/.*)?)
nodetypeStringJCR node type the event node must be (e.g., dam:AssetContent)
conditionsString[]Additional JCR property conditions on the event node
workflowStringRuntime path of the workflow model /var/workflow/models/<id>
enabledBooleanWhether the launcher is active
descriptionStringHuman-readable description
excludeListStringComma-separated list of entries that suppress the launcher for matching events. Two entry formats can be mixed: bare JCR property names (skip events whose only changed property matches, e.g. jcr:lastModified) and event-user-data:<value> (skip events tagged with that JCR observation userData, e.g. event-user-data:changedByWorkflowProcess).
runModesString[]Restrict to specific run modes — honoring is unreliable on 6.5 LTS; prefer config.author/ packaging
transientBooleanRun the launched workflow as transient — no /var/workflow/instances/ node unless persistence is forced. Use for high-volume launchers
noProcessBooleanMatch events but do not start the workflow — silence a launcher without removing it

Launcher Storage Paths on 6.5 LTS

On AEM 6.5 LTS, launcher configurations can live at:

PathNotes
/libs/settings/workflow/launcher/config/OOTB launchers — do not edit directly
/conf/global/settings/workflow/launcher/config/Recommended for new custom launchers
/apps/settings/workflow/launcher/config/Alternative overlay location
/etc/workflow/launcher/config/Legacy path (AEM 6.0–6.2); still supported but migrate away

Resolution order: /conf/global → /apps → /libs

Deploying a Custom Launcher on 6.5 LTS

Maven project location:

ui.content/src/main/content/jcr_root/conf/global/settings/workflow/launcher/config/
    my-custom-launcher/
        .content.xml

Or for overlay-based approach under /apps:

ui.apps/src/main/content/jcr_root/apps/settings/workflow/launcher/config/
    my-custom-launcher/
        .content.xml

Node structure (.content.xml):

xml
<?xml version="1.0" encoding="UTF-8"?>
<jcr:root
    xmlns:jcr="http://www.jcp.org/jcr/1.0"
    xmlns:cq="http://www.day.com/jcr/cq/1.0"
    jcr:primaryType="cq:WorkflowLauncher"
    eventType="{Long}1"
    glob="/content/dam(/.*)?/jcr:content/renditions/original"
    nodetype="nt:file"
    workflow="/var/workflow/models/dam/update_asset"
    enabled="{Boolean}true"
    description="Start DAM update workflow on new original rendition upload"/>
<!-- For author-only restriction, package this .content.xml under config.author/.
     The runModes property on cq:WorkflowLauncher is unreliable on 6.5 LTS. -->

Filter in filter.xml:

xml
<filter root="/conf/global/settings/workflow/launcher/config/my-custom-launcher"/>

Overlaying an OOTB Launcher

To disable or modify an OOTB launcher:

  1. Copy the node from /libs/settings/workflow/launcher/config/<launcher-name> to /conf/global/settings/workflow/launcher/config/<launcher-name> (or /apps/settings/...)
  2. Modify the property (e.g., enabled="{Boolean}false")
  3. Deploy via Package Manager or Maven

Common OOTB Launchers (6.5 LTS)

Verified against /libs/settings/workflow/launcher/config/ on AEM 6.5 LTS. Inspect any launcher in CRXDE Lite or via Tools → Workflow → Launchers for the full property set.

LaunchereventTypenodetypeglobworkflow
update_asset_create1 (NODE_ADDED)nt:file/content/dam(/((?!/subassets).)*/)renditions/original/var/workflow/models/dam/update_asset
update_asset_mod16 (PROPERTY_CHANGED)nt:file/content/dam(/((?!/subassets).)*/)renditions/original/var/workflow/models/dam/update_asset
dam_xmp_writeback16 (PROPERTY_CHANGED)nt:unstructured/content/dam(/.*)/jcr:content/metadata/var/workflow/models/dam-xmp-writeback

Event Type Combinations

To listen for both ADD and MODIFY, combine event types:

xml
eventType="{Long}3"  <!-- 1 (ADD) + 2 (MODIFY) = 3 -->

Where-Clause Conditions

xml
conditions="[property=cq:type,value=publicationevent,type=STRING]"

Condition format: property=<name>,value=<value>,type=<JCR_TYPE> (type is optional, defaults to STRING).

Architecture Considerations

Launchers are the surface where workflow load is automated. A single bad glob can flood the workflow job queue with one content edit. Apply these before deploying any launcher:

  • Narrow your glob, node type, and conditions. A broad glob (/content(/.*)?) paired with eventType=2 (NODE_MODIFIED) fires on every property change under /content. Always pair the glob with a specific nodetype (cq:PageContent, dam:AssetContent) and conditions that match only the events you care about.
  • Watch multi-event amplification. A single DAM asset upload fires multiple events — the asset node, each rendition, the metadata node. Without narrowing, one upload starts N workflows. Pin the glob to a specific descendant (e.g., /jcr:content/renditions/original) when you only want one trigger per upload.
  • Avoid infinite loops. A workflow whose process step writes to a path the launcher watches will re-trigger itself. Default strategy: tag the JCR Session with setUserData("changedByWorkflowProcess") before the write so WorkflowLauncherListener ignores the resulting events. The session parameter on WorkflowProcess.execute() is a WorkflowSession, not a JCR Session — adapt it first: javax.jcr.Session jcrSession = session.adaptTo(javax.jcr.Session.class); jcrSession.getWorkspace().getObservationManager().setUserData("changedByWorkflowProcess");. If you write through a service-user ResourceResolver instead, tag that resolver's underlying Session — it is a different Session instance and the flag does not propagate. This is the most robust pattern (works across model changes, no static config to keep in sync). Use the launcher's excludeList only when you can statically name every model that might re-trigger; use a JCR property flag when the workflow writes to a different node than the launcher watches. See condition-patterns.md for code.
  • Use transient workflows for high-volume launchers. Set transient="true" on the workflow model (see workflow-model-design Architecture Considerations). Persistent workflows in this regime bloat /var/workflow/instances quickly.
  • Disable broad launchers in lower environments. A broad-match OOTB or custom launcher active in dev/stage with the same content as prod can fire on every content sync, masking real prod-vs-dev behavior. Either disable in lower envs (enabled={Boolean}false overlay), package the launcher under config.author/ for run-mode restriction, or restrict with conditions tied to a prod-only marker.
  • Stack mechanisms cautiously. Pairing a Workflow Launcher with a Replication Trigger on the same content fires two workflows per content change. Pick one mechanism per content event.

Debugging Launchers (6.5 LTS)

Local development only. The curl -u admin:admin example below targets a local author at localhost:4502 with the default admin password. Never run it against a shared, stage, or production instance, and never with default admin:admin credentials outside an isolated dev box.

  • Tools → Workflow → Launchers UI — lists all active launchers, interactive enable/disable
  • Check /conf/global/settings/workflow/launcher/config/ and /apps/settings/workflow/launcher/config/ in CRXDE Lite
  • Felix Web Console → OSGi → WorkflowLauncherListener service
  • Check /var/workflow/launcher/ for active event registrations
  • Run curl -u admin:admin http://localhost:4502/etc/workflow/launcher.json to list all

Verifying the Launcher

After deploying a launcher, confirm it's loaded and firing:

SurfaceHow
Author UITools → Workflow → Launchers — your launcher should appear with the Enabled flag set
CRXDEConfirm the node exists at /conf/global/settings/workflow/launcher/config/<name> (or /apps/) with all properties intact
Trigger testManually create the event (upload a file to the watched path; edit a page) and check Tools → Workflow → Instances for a new instance
LogsEnable DEBUG for com.adobe.granite.workflow.core.launcher in the Felix console — every launched instance produces a log line

When the launcher doesn't fire

If the event happens but no instance appears:

  • Most common: the glob doesn't match the actual event node path. The event for a DAM asset upload fires on the dam:AssetContent node (/content/dam/.../jcr:content), not on dam:Asset itself. Inspect the event node path in CRXDE before tuning the glob.
  • Node type mismatch: the nodetype= doesn't match the event node's primary type. Confirm in CRXDE.
  • Path resolution: a same-named launcher node at higher-priority /conf/global may shadow the one you deployed at /apps/. Check the resolution order in Launcher Storage Paths above.
  • Globally excluded path: events under specific paths are ignored regardless of launcher config — see condition-patterns.md.

If the launcher fires (you see the log line) but the workflow doesn't progress, the issue is downstream — most commonly a referenced WorkflowProcess is not registered (see workflow-development Dependencies). For full diagnosis: workflow-debugging.

References in This Skill

ReferenceWhat It Covers
references/workflow-launchers/launcher-config-reference.mdFull property spec and XML templates
references/workflow-launchers/condition-patterns.mdCommon condition patterns, glob syntax, event type codes
../workflow-orchestrator/references/workflow-foundation/architecture-overview.mdGranite Workflow Engine overview (canonical)
../workflow-orchestrator/references/workflow-foundation/65-lts-guardrails.md6.5 LTS constraints and legacy path guidance (canonical)
../workflow-orchestrator/references/workflow-foundation/jcr-paths-reference.mdWhere launchers live in the JCR (canonical)

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 Launchers AI skill do?

Configure and deploy Workflow Launchers that automatically start workflows in response to JCR content changes on AEM 6.5 LTS

Why use Workflow Launchers on TypingMind?

Because you install it once and use it with any model. Workflow Launchers 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 Launchers 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-launchers. 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 Launchers?

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 Launchers?

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

Is the Workflow Launchers 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 👇