القائمة
Instructions on building, maintaining, and extending the Liferay DXP build-time configuration compatibility transformations (create-fragment-zips.sh) using jq.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Activate this skill whenever you are running E2E tests, capturing screenshots, promoting visual baselines, or interpreting E2E test results for Liferay fragments. It specifies the exact commands for full-suite runs, targeted fragment runs, and visual baseline promotion, and documents the current known platform blocker.
Activate this skill when a task is large enough to benefit from splitting work across concurrent or hierarchical subagents — for example, running research and implementation in parallel, or delegating lint/test execution to a background agent while the coordinator continues planning. It defines the collaboration patterns, spawn constraints, output integration rules, and conflict resolution protocol.
Activate this skill when executing any multi-phase fragment development task — scaffolding, linting, compatibility build, E2E testing, or documentation update — where the output of one phase is the required input to the next. It defines the canonical pipeline phases, their inputs/outputs, and the active constraints that prevent a downstream phase from starting before its upstream gate has passed.
Activate this skill whenever you are creating a branch, opening a Pull Request, merging code, or performing any Git / GitHub CLI operation in the liferay-fragments repository. It enforces branch naming conventions, pre-commit validation gates, the PR auto-merge protocol, and the code-review requirement.
Activate this skill whenever you are planning a new feature, bug fix, enhancement, or any tracked piece of work. It mandates issue-first development and prescribes the exact workflow for creating, syncing, and closing GitHub Issues using the gh-issue-sync.cjs automation tool.
Activate this skill whenever you are building, packaging, or releasing fragment ZIP collections. It defines the mandatory three-target build strategy that produces separate ZIPs for Liferay 2026.Q1+, pre-2026.Q1, and pre-2025.Q3, and specifies the exact dataType and checkbox default-value transformation rules for each target.
| name | liferay-compat-transform |
| description | Instructions on building, maintaining, and extending the Liferay DXP build-time configuration compatibility transformations (create-fragment-zips.sh) using jq. |
This skill allows the agent to build, maintain, and extend the fragment configuration transformation pipeline (create-fragment-zips.sh) to support different Liferay DXP releases.
Liferay DXP validates configuration.json properties (dataType, defaultValue, dependencies) differently across versions:
"dataType": "number" on numeric fields and strict boolean literals for checkbox defaults."dataType": "int" instead of "number", deletes conditional dependencies, and expects string defaults everywhere.To avoid duplicate code, we maintain the single modern schema source in the repository and run a pipeline in create-fragment-zips.sh using jq to build target ZIPs.
If a future Liferay release (e.g., 2027.Q1) introduces a breaking change to the configuration metadata requirements:
Define a Build Directory:
In create-fragment-zips.sh, copy the collection source into a new temp directory:
BUILD_TEMP_2027="temp_post2027q1_${COLLECTION_NAME}"
mkdir -p "$BUILD_TEMP_2027/$COLLECTION_NAME"
cp -r "$COLLECTION_NAME/." "$BUILD_TEMP_2027/$COLLECTION_NAME/"
Apply the jq Transformation:
Transform configuration.json using a jq filter:
find "$BUILD_TEMP_2027/$COLLECTION_NAME" -name "configuration.json" -print0 | while IFS= read -r -d '' config_file; do
jq "<YOUR_JQ_FILTER>" "$config_file" > "$config_file.tmp" && mv "$config_file.tmp" "$config_file"
done
Common JQ Recipes:
"(.. | objects | select(.oldKey != null)) .newKey = .oldKey | del(.oldKey)""(.. | objects | select(.dataType == \"number\" and .defaultValue != null)) .defaultValue |= tostring""del(.fieldSets[].fields[].typeOptions.dependency)"Zip and Clean up: Zip the directory using the new suffix target:
OUTPUT_ZIP_2027_ABS="$ZIPS_ROOT/fragments/${COLLECTION_NAME}-post2027q1${BUILD_SUFFIX}.zip"
(cd "$BUILD_TEMP_2027" && zip -qr "../$OUTPUT_ZIP_2027_ABS" . -x "*.DS_Store")
rm -rf "$BUILD_TEMP_2027"
Verify Locally: Run the local E2E suite targeting the matching tag:
$env:PORT="8081"; & "C:\Program Files\Git\bin\bash.exe" scripts/test-runner.sh -p e2e-test-env -f <collection-name>