Respond To Coderabbitai

Operational recipe

1) Discover all unresolved review comments (GraphQL - Recommended)

RECOMMENDED: Use GraphQL to get unresolved review threads directly. This is more reliable than filtering REST API results.

REPO="stonelyd/altium-design-review"  # Update with your repo
PR=5                                    # Update with your PR number

# Get all unresolved review threads with full comment details
readarray -t THREADS < <(
gh api graphql -f query='
query($owner:String!, $name:String!, $pr:Int!) {
  repository(owner:$owner, name:$name) {
    pullRequest(number:$pr) {
      reviewThreads(first: 100) {
        nodes {
          id
          isResolved
          comments(first: 50) {
            nodes {
              databaseId
              path
              originalPosition
              diffHunk
              isMinimized
              author { login }
            }
          }
        }
      }
    }
  }
}' \
-F owner="${REPO%%/*}" -F name="${REPO##*/}" -F pr="$PR" \
| jq -c '.data.repository.pullRequest.reviewThreads.nodes[]
| select(.isResolved==false)
| {thread_id:.id,
   top_comment_id:(.comments.nodes[0].databaseId),
   path:(.comments.nodes[0].path),
   author:(.comments.nodes[0].author.login),
   diff:(.comments.nodes[0].diffHunk)}')

# Display the threads
echo "=== Found ${#THREADS[@]} unresolved threads ==="
for i in "${!THREADS[@]}"; do
  echo -e "\n--- Thread $((i+1)) ---"
  echo "${THREADS[$i]}" | jq -r '"Comment ID: \(.top_comment_id)\nFile: \(.path)\nAuthor: \(.author)\nThread ID: \(.thread_id)"'
done

Alternative: REST API for comment bodies

Use REST API to get full comment bodies after you have the comment IDs from GraphQL.

# Save all comments to file for processing (includes threaded replies)
gh api repos/${REPO}/pulls/${PR}/comments --paginate > /tmp/all_comments.json

# Extract specific comment details
for CID in 2466210587 2466210599 2466210608; do
  echo -e "\n=========================================="
  echo "COMMENT ID: $CID"
  cat /tmp/all_comments.json | jq -r ".[] | select(.id == $CID) | \"File: \(.path):\(.line // .original_line // \"null\")\\nAuthor: \(.user.login)\\n\\nBody:\\n\(.body)\\n\""
  echo "=========================================="
done

2) Get full comment bodies for each unresolved thread

After getting thread IDs from GraphQL, fetch full comment bodies from REST API:

# Method 1: Batch fetch all comments to file (recommended for many comments)
gh api repos/${REPO}/pulls/${PR}/comments --paginate > /tmp/all_comments.json

# Extract full body for specific comment IDs
for CID in "${COMMENT_IDS[@]}"; do
  echo -e "\n=========================================="
  echo "COMMENT ID: $CID"
  cat /tmp/all_comments.json | jq -r ".[] | select(.id == $CID) | \"File: \(.path):\(.line // \"null\")\\nAuthor: \(.user.login)\\n\\nBody:\\n\(.body)\\n\""
  echo "=========================================="
done

# Method 2: Fetch individual comment (for single/few comments)
CID=2466210587
gh api repos/${REPO}/pulls/${PR}/comments/${CID} | jq -r '"File: \(.path):\(.line // \"null\")\nAuthor: \(.user.login)\n\nBody:\n\(.body)"'

Key learnings:

• GraphQL databaseId matches REST API id field
• GraphQL doesn't include full comment body text - use REST API for that
• Always verify in_reply_to_id is null (GraphQL filters this automatically via reviewThreads)

3) Implement the fix

• Open the referenced file/section.
• Apply the minimal change that satisfies the reviewer's request.
• MANDATORY: Run all tests before committing:
  1. npm run check - TypeScript type checking and linting
  2. npm run test - Unit tests
  3. npm run test:components - Component tests (if changes affect UI components)
• Fix any test failures or type errors before proceeding to commit.

4) Create atomic commits per logical issue

CRITICAL: Create one commit per logical issue/fix. Group related comments that address the same issue into a single atomic commit.

Commit strategy:

• Single issue = Single commit: If multiple comments point to the same underlying problem, fix it in one commit
• Different issues = Separate commits: Keep unrelated fixes in separate commits for clean history
• Atomic changes: Each commit should be self-contained and make sense on its own

Examples:

• ✅ Good: One commit fixing "Add archived filter" that addresses 3 comments about missing archived filters in different functions
• ✅ Good: Separate commits for "Add dimension validation", "Remove unused variable", "Fix exit code handling"
• ❌ Bad: One mega-commit fixing all 15 unrelated review comments
• ❌ Bad: 15 tiny commits when 3 comments are about the same validation issue

FILE="path/from/context"
SUMMARY="Add archived=false filters to match functions"

git add "$FILE"
COMMIT_MSG=$(cat <<EOF