name: Index Artifact Management
on:
workflow_dispatch:
inputs:
action:
description: 'Action to perform'
required: true
default: 'validate'
type: choice
options:
- validate
- promote
- cleanup
- list
artifact_name:
description: 'Artifact name (for promote/validate)'
required: false
type: string
retention_days:
description: 'Retention days for promoted artifacts'
required: false
default: '30'
type: string
workflow_call:
inputs:
index_path:
description: 'Path to index files'
required: true
type: string
artifact_suffix:
description: 'Suffix for artifact name'
required: false
default: ''
type: string
retention_days:
description: 'Days to retain artifact'
required: false
default: '7'
type: number
secrets:
GITHUB_TOKEN:
required: true
env:
COMPRESSION_LEVEL: '9'
MAX_ARTIFACT_SIZE_MB: '500'
INDEX_FILES: |
code_index.db
vector_index.qdrant/
.index_metadata.json
jobs:
upload-index-artifact:
name: Upload Index Artifact
runs-on: ubuntu-latest
if: github.event_name == 'workflow_call'
outputs:
artifact_name: ${{ steps.upload.outputs.artifact_name }}
artifact_size: ${{ steps.compress.outputs.size }}
checksum: ${{ steps.compress.outputs.checksum }}
steps:
- name: Checkout code
uses: actions/checkout@v4
- name: Prepare artifact metadata
id: metadata
run: |
# Generate artifact name
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
COMMIT_SHORT=$(git rev-parse --short HEAD)
SUFFIX="${{ inputs.artifact_suffix }}"
if [ -n "$SUFFIX" ]; then
ARTIFACT_NAME="index-${COMMIT_SHORT}-${SUFFIX}-${TIMESTAMP}"
else
ARTIFACT_NAME="index-${COMMIT_SHORT}-${TIMESTAMP}"
fi
echo "artifact_name=$ARTIFACT_NAME" >> $GITHUB_OUTPUT
# Get index stats
if [ -f "${{ inputs.index_path }}/code_index.db" ]; then
DB_SIZE=$(stat -c%s "${{ inputs.index_path }}/code_index.db" 2>/dev/null || stat -f%z "${{ inputs.index_path }}/code_index.db" 2>/dev/null || echo "0")
echo "db_size=$DB_SIZE" >> $GITHUB_OUTPUT
fi
- name: Compress index files
id: compress
run: |
cd "${{ inputs.index_path }}"
# Create compressed archive
echo "Creating compressed archive..."
tar -czf index-archive.tar.gz \
--exclude='*.log' \
--exclude='*.tmp' \
${{ env.INDEX_FILES }} 2>/dev/null || true
# Calculate checksum
CHECKSUM=$(sha256sum index-archive.tar.gz | cut -d' ' -f1)
echo "checksum=$CHECKSUM" >> $GITHUB_OUTPUT
# Get compressed size
SIZE=$(stat -c%s index-archive.tar.gz 2>/dev/null || stat -f%z index-archive.tar.gz 2>/dev/null)
SIZE_MB=$((SIZE / 1024 / 1024))
echo "size=$SIZE_MB MB" >> $GITHUB_OUTPUT
# Check size limit
if [ $SIZE_MB -gt ${{ env.MAX_ARTIFACT_SIZE_MB }} ]; then
echo "❌ Compressed archive too large: ${SIZE_MB}MB > ${MAX_ARTIFACT_SIZE_MB}MB"
exit 1
fi
# Create metadata file
cat > artifact-metadata.json <<EOF
{
"artifact_name": "${{ steps.metadata.outputs.artifact_name }}",
"commit": "$(git rev-parse HEAD)",
"branch": "$(git branch --show-current)",
"timestamp": "$(date -u +%Y-%m-%dT%H:%M:%SZ)",
"compressed_size": $SIZE,
"checksum": "$CHECKSUM",
"files_included": $(echo '${{ env.INDEX_FILES }}' | jq -R -s -c 'split("\n") | map(select(length > 0))'),
"compression_level": ${{ env.COMPRESSION_LEVEL }},
"index_stats": {
"db_size": ${{ steps.metadata.outputs.db_size || 0 }}
}
}
EOF
- name: Upload artifact
id: upload
uses: actions/upload-artifact@v4
with:
name: ${{ steps.metadata.outputs.artifact_name }}
path: |
${{ inputs.index_path }}/index-archive.tar.gz
${{ inputs.index_path }}/artifact-metadata.json
retention-days: ${{ inputs.retention_days }}
compression-level: 0 # Already compressed
- name: Output summary
run: |
echo "### 📦 Index Artifact Uploaded" >> $GITHUB_STEP_SUMMARY
echo "" >> $GITHUB_STEP_SUMMARY
echo "- **Name:** \`${{ steps.metadata.outputs.artifact_name }}\`" >> $GITHUB_STEP_SUMMARY
echo "- **Size:** ${{ steps.compress.outputs.size }}" >> $GITHUB_STEP_SUMMARY
echo "- **Checksum:** \`${{ steps.compress.outputs.checksum }}\`" >> $GITHUB_STEP_SUMMARY
echo "- **Retention:** ${{ inputs.retention_days }} days" >> $GITHUB_STEP_SUMMARY
validate-artifact:
name: Validate Index Artifact
runs-on: ubuntu-latest
if: github.event_name == 'workflow_dispatch' && inputs.action == 'validate'
steps:
- name: Checkout code
uses: actions/checkout@v4
- name: Download artifact
uses: actions/download-artifact@v4
with:
name: ${{ inputs.artifact_name }}
path: ./artifact-download
- name: Validate artifact
run: |
cd artifact-download
# Check files exist
if [ ! -f "index-archive.tar.gz" ] || [ ! -f "artifact-metadata.json" ]; then
echo "❌ Missing required files"
exit 1
fi
# Read metadata
EXPECTED_CHECKSUM=$(jq -r '.checksum' artifact-metadata.json)
ACTUAL_CHECKSUM=$(sha256sum index-archive.tar.gz | cut -d' ' -f1)
# Verify checksum
if [ "$EXPECTED_CHECKSUM" != "$ACTUAL_CHECKSUM" ]; then
echo "❌ Checksum mismatch!"
echo "Expected: $EXPECTED_CHECKSUM"
echo "Actual: $ACTUAL_CHECKSUM"
exit 1
fi
# Extract and validate contents
tar -tzf index-archive.tar.gz > /dev/null
echo "✅ Artifact validated successfully"
# Display metadata
echo "### 📋 Artifact Metadata" >> $GITHUB_STEP_SUMMARY
jq . artifact-metadata.json >> $GITHUB_STEP_SUMMARY
promote-artifact:
name: Promote Artifact
runs-on: ubuntu-latest
if: github.event_name == 'workflow_dispatch' && inputs.action == 'promote'
steps:
- name: Download artifact
uses: actions/download-artifact@v4
with:
name: ${{ inputs.artifact_name }}
path: ./artifact-download
- name: Re-upload with extended retention
uses: actions/upload-artifact@v4
with:
name: ${{ inputs.artifact_name }}-promoted
path: ./artifact-download/*
retention-days: ${{ inputs.retention_days }}
- name: Summary
run: |
echo "### ✅ Artifact Promoted" >> $GITHUB_STEP_SUMMARY
echo "" >> $GITHUB_STEP_SUMMARY
echo "- **Original:** \`${{ inputs.artifact_name }}\`" >> $GITHUB_STEP_SUMMARY
echo "- **Promoted:** \`${{ inputs.artifact_name }}-promoted\`" >> $GITHUB_STEP_SUMMARY
echo "- **New Retention:** ${{ inputs.retention_days }} days" >> $GITHUB_STEP_SUMMARY
cleanup-artifacts:
name: Cleanup Old Artifacts
runs-on: ubuntu-latest
if: github.event_name == 'workflow_dispatch' && inputs.action == 'cleanup' || github.event_name == 'schedule'
steps:
- name: Setup Python
uses: actions/setup-python@v5
with:
python-version: '3.11'
- name: List and cleanup artifacts
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
run: |
pip install PyGithub
python << 'EOF'
import os
from datetime import datetime, timedelta
from github import Github
g = Github(os.environ['GITHUB_TOKEN'])
repo = g.get_repo(os.environ['GITHUB_REPOSITORY'])
# Configuration
MAX_AGE_DAYS = 30
KEEP_LATEST_COUNT = 5
PROMOTED_SUFFIX = '-promoted'
cutoff_date = datetime.now() - timedelta(days=MAX_AGE_DAYS)
# Get all artifacts
artifacts = list(repo.get_artifacts())
index_artifacts = [a for a in artifacts if a.name.startswith('index-')]
# Separate promoted and regular artifacts
promoted = [a for a in index_artifacts if PROMOTED_SUFFIX in a.name]
regular = [a for a in index_artifacts if PROMOTED_SUFFIX not in a.name]
# Sort by created date (newest first)
regular.sort(key=lambda x: x.created_at, reverse=True)
# Keep latest N artifacts
to_delete = []
# Delete old regular artifacts (keep latest N)
for i, artifact in enumerate(regular):
if i >= KEEP_LATEST_COUNT and artifact.created_at < cutoff_date:
to_delete.append(artifact)
# Delete very old promoted artifacts
for artifact in promoted:
if artifact.created_at < cutoff_date:
to_delete.append(artifact)
# Perform deletion
print(f"Found {len(index_artifacts)} index artifacts")
print(f"Will delete {len(to_delete)} old artifacts")
for artifact in to_delete:
print(f"Deleting: {artifact.name} (created: {artifact.created_at})")
artifact.delete()
print(f"\nCleanup complete. Remaining artifacts: {len(index_artifacts) - len(to_delete)}")
EOF
list-artifacts:
name: List Available Artifacts
runs-on: ubuntu-latest
if: github.event_name == 'workflow_dispatch' && inputs.action == 'list'
steps:
- name: Setup Python
uses: actions/setup-python@v5
with:
python-version: '3.11'
- name: List artifacts
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
run: |
pip install PyGithub tabulate
python << 'EOF'
import os
from datetime import datetime
from github import Github
from tabulate import tabulate
g = Github(os.environ['GITHUB_TOKEN'])
repo = g.get_repo(os.environ['GITHUB_REPOSITORY'])
# Get all index artifacts
artifacts = list(repo.get_artifacts())
index_artifacts = [a for a in artifacts if a.name.startswith('index-')]
# Sort by created date (newest first)
index_artifacts.sort(key=lambda x: x.created_at, reverse=True)
# Prepare table data
table_data = []
for artifact in index_artifacts[:20]: # Show latest 20
age_days = (datetime.now() - artifact.created_at.replace(tzinfo=None)).days
size_mb = artifact.size_in_bytes / 1024 / 1024
promoted = "Yes" if "-promoted" in artifact.name else "No"
table_data.append([
artifact.name,
f"{size_mb:.1f} MB",
artifact.created_at.strftime("%Y-%m-%d %H:%M"),
f"{age_days} days",
promoted
])
# Print table
print("## 📦 Available Index Artifacts\n")
headers = ["Name", "Size", "Created", "Age", "Promoted"]
print(tabulate(table_data, headers=headers, tablefmt="pipe"))
# Summary
print(f"\nTotal artifacts: {len(index_artifacts)}")
total_size_mb = sum(a.size_in_bytes for a in index_artifacts) / 1024 / 1024
print(f"Total size: {total_size_mb:.1f} MB")
EOF
