Agent skill
test-traditional-build
Run a traditional AIB build test and report results. Builds a disk image using automotive-image-builder manifest and downloads the artifact.
Stars
163
Forks
31
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/development/test-traditional-build-centos-automotive-su-automotive-dev-opera
SKILL.md
Traditional Build Test Workflow
Run a traditional AIB build with the specified parameters, monitor until completion, and download the artifact.
Parameters
$ARGUMENTS contains the build name (required)
Required Environment Variables
Before running, ensure these are set:
PUSH_REGISTRY- Registry URL to push the disk image (e.g.,quay.io/myorg/traditional-disk)REGISTRY_USERNAME- Registry username (or read fromregistry-userfile)REGISTRY_PASSWORD- Registry password (or read fromregistry-passfile)
Execution Steps
1. Validate Environment
bash
# Check required environment
if [ -z "$PUSH_REGISTRY" ]; then
echo "ERROR: PUSH_REGISTRY environment variable not set"
echo "Example: export PUSH_REGISTRY=quay.io/myorg/traditional-disk"
exit 1
fi
2. Infer CAIB_SERVER from OpenShift Route
bash
CAIB_SERVER="https://$(oc get route ado-build-api -n automotive-dev-operator-system -o jsonpath='{.spec.host}')"
echo "Using CAIB_SERVER: $CAIB_SERVER"
3. Get Authentication Token
bash
TOKEN=$(oc whoami -t)
4. Get Registry Credentials
bash
# Try environment variables first, fall back to files
REGISTRY_USERNAME="${REGISTRY_USERNAME:-$(cat registry-user 2>/dev/null)}"
REGISTRY_PASSWORD="${REGISTRY_PASSWORD:-$(cat registry-pass 2>/dev/null)}"
5. Submit Build with Download
bash
BUILD_NAME="$ARGUMENTS"
PUSH_IMAGE="${PUSH_REGISTRY}:latest"
OUTPUT_DIR="./output"
bin/caib build-traditional simple.aib.yml \
--server "$CAIB_SERVER" \
--token "$TOKEN" \
--name "$BUILD_NAME" \
--arch arm64 \
--target qemu \
--export qcow2 \
--compress gzip \
--push "$PUSH_IMAGE" \
--registry-username "$REGISTRY_USERNAME" \
--registry-password "$REGISTRY_PASSWORD" \
--wait \
--follow \
--download "$OUTPUT_DIR"
6. Monitor Build (if not using --follow)
bash
# Check build status
oc get imagebuild -n automotive-dev-operator-system $BUILD_NAME -o jsonpath='{.status.phase}'
7. On Failure - Debug
bash
# Get PipelineRun name from ImageBuild status
PIPELINE_RUN=$(oc get imagebuild -n automotive-dev-operator-system $BUILD_NAME -o jsonpath='{.status.pipelineRunName}')
# Get pod name using correct PipelineRun label
POD=$(oc get pods -n automotive-dev-operator-system -l tekton.dev/pipelineRun=${PIPELINE_RUN} -o jsonpath='{.items[0].metadata.name}')
# Get build logs
oc logs -n automotive-dev-operator-system $POD -c step-build-image --tail=200
# Check for common issues
oc logs -n automotive-dev-operator-system $POD -c step-build-image | grep -i "error\|failed\|unauthorized"
8. Verify Download
bash
# Check if artifact was downloaded
OUTPUT_FILE=$(ls -1 ${OUTPUT_DIR}/*.qcow2* 2>/dev/null | head -1)
if [ -n "$OUTPUT_FILE" ] && [ -f "$OUTPUT_FILE" ]; then
echo "SUCCESS: Artifact downloaded to $OUTPUT_FILE"
ls -lh "$OUTPUT_FILE"
else
echo "WARNING: Artifact not found in $OUTPUT_DIR"
ls -la "$OUTPUT_DIR" 2>/dev/null || echo "Output directory does not exist"
fi
9. Report Results
Provide a summary with:
- Build status (success/failure)
- Download location and file size
- If failed: the specific error and relevant log snippets
- Whether logs streamed correctly (testing the pipelineRun label fix)
- Suggestions for fixes based on the error type
Common Error Patterns
- "log stream not ready (HTTP 503)" - Pod label selector issue (should use
tekton.dev/pipelineRun) - "unauthorized" - Token or registry auth issues
- "PUSH_REGISTRY not set" - Environment variable missing
- "manifest not found" - AIB manifest file missing or invalid
Alternative: Download Only (for completed builds)
bash
bin/caib download \
--server "$CAIB_SERVER" \
--token "$TOKEN" \
--name "$BUILD_NAME" \
--output-dir ./output
Didn't find tool you were looking for?