Extract the information about images in your harbor, so you can identify images, which haven't been used and get rid of them

generate_harbor_report.sh

#!/bin/bash

# Harbor Configuration
HARBOR_URL="https://myharbor.com"
HARBOR_USER="my_robot_user"
HARBOR_PASS="my_robot_users_password"

OUTPUT_CSV="harbor_images.csv"
SORTED_CSV="sorted_harbor_images.csv"
LOG_FILE="harbor_debug.log"

# Initialize files
echo "project,repository,tag,last_pull_time,push_time,size_bytes,digest" > "$OUTPUT_CSV"
echo "Debug Log - $(date)" > "$LOG_FILE"

# Fetch all projects
projects=$(curl -s -u "${HARBOR_USER}:${HARBOR_PASS}" \
  "${HARBOR_URL}/api/v2.0/projects?page_size=100" \
  | jq -r '.[].name')

# Loop through each project
for PROJECT in $projects; do

  # Fetch all repositories for the project
  repos=$(curl -s -u "${HARBOR_USER}:${HARBOR_PASS}" \
    "${HARBOR_URL}/api/v2.0/projects/${PROJECT}/repositories?page_size=100" \
    | jq -r '.[].name' | sed "s|^${PROJECT}/||")

  # Loop through each repository
  for REPO in $repos; do

    # Fetch artifact details
    RESPONSE=$(curl -s -u "${HARBOR_USER}:${HARBOR_PASS}" \
      "${HARBOR_URL}/api/v2.0/projects/${PROJECT}/repositories/${REPO}/artifacts?page_size=100")

    # Check for errors in response
    if echo "$RESPONSE" | jq -e '.errors' > /dev/null; then
      echo "API Error for project: $PROJECT, repo: $REPO" >> "$LOG_FILE"
      echo "$RESPONSE" >> "$LOG_FILE"
      echo "----" >> "$LOG_FILE"
      continue
    fi

    # Parse JSON safely with additional fields
    echo "$RESPONSE" | jq -r --arg PROJECT "$PROJECT" --arg REPO "$REPO" '
      .[] | [
        $PROJECT,
        $REPO,
        (.tags // [] | if length > 0 then .[0].name else "untagged" end),
        (.pull_time // "1970-01-01T00:00:00Z"),
        (.push_time // "1970-01-01T00:00:00Z"),
        (.size // 0),
        (.digest // "N/A")
      ] | @csv
    ' >> "$OUTPUT_CSV" 2>> "$LOG_FILE"

  done
done

# Sort CSV by last_pull_time (newest first)
sort -t, -k4 -r "$OUTPUT_CSV" > "$SORTED_CSV"

# Output result
echo "✅ Report successfully generated: $SORTED_CSV"
echo "📌 Check '$LOG_FILE' for any remaining issues."