🚀 How to Retrieve Stale Branches Across All Repositories in a GitHub Organization
Keeping your GitHub repositories clean is crucial for maintainability. Over time, branches pile up, making it difficult to track what’s relevant. If you’re managing multiple repositories in an organization, manually identifying stale branches can be a hassle.
Luckily, with the GitHub CLI (gh), you can automate this process and generate a report of all non-main branches along with their last authors.
This guide walks you through a Bash script that:
✅ Retrieves all repositories in an organization
✅ Lists branches, excluding main, master, and azure-master
✅ Identifies the last commit author for each branch
✅ Groups the results by author for better visibility
🔧 Prerequisites
Before running the script, make sure:
- You have GitHub CLI (
gh) installed and authenticated (gh auth login). - You have appropriate permissions to list repositories and fetch branch details in your organization.
- You’re running a Unix-based system (Linux/macOS or WSL for Windows).
📜 The Script
Here’s the complete Bash script to scan all repositories in your GitHub organization and generate a stale branch report:
# Set Variables
org="<Replace with your GitHub Organization name>"
raw_output_file="branches.txt"
report_file="report.txt"
# Clear previous output
> "$raw_output_file"
# Get the repositories as JSON and extract names
repos=$(gh repo list "$org" --json name --jq '.[].name')
# Function to get branch author safely
get_author() {
local org=$1
local repo=$2
local branch=$3
local author
# Try fetching author name
author=$(gh api "/repos/$org/$repo/branches/$branch" --jq '.commit.commit.author.name' 2>/dev/null)
# If author is empty, return "Unknown"
if [[ -z "$author" ]]; then
author="Unknown"
fi
# Write to output file
echo " - Author: $author" >> "$raw_output_file"
}
# Loop through each repository
for repo in $repos; do
echo "- repo: $repo" >> "$raw_output_file"
# Get all branches in the repository
branches=$(gh api "/repos/$org/$repo/branches" --jq '.[].name')
# Loop through each branch
for branch in $branches; do
# Skip master, azure-master, and main
if [[ "$branch" == "master" || "$branch" == "azure-master" || "$branch" == "main" ]]; then
continue
fi
echo " - Branch: $branch" >> "$raw_output_file"
# Fetch branch author
get_author "$org" "$repo" "$branch"
done
done
declare -A author_repos # Declare an associative array
# Ensure the array is cleared before each run
unset author_repos
declare -A author_repos
# Clear previous report
> "$report_file"
# Read and process output file
while IFS= read -r line; do
if [[ $line == "- repo: "* ]]; then
repo=$(echo "$line" | awk '{print $3}')
elif [[ $line == " - Branch: "* ]]; then
branch=$(echo "$line" | awk '{print $3}')
elif [[ $line == " - Author: "* ]]; then
author=$(echo "$line" | sed 's/ - Author: //') # Extract author name
# Store branch under author
author_repos["$author"]+=$'\n'"- $repo -> $branch"
fi
done < "$raw_output_file"
# Print grouped results
for author in "${!author_repos[@]}"; do
echo "Author: $author" >> "$report_file"
echo "${author_repos[$author]}" >> "$report_file"
echo >> "$report_file"
done
cat "$report_file"
📊 Example Output
Once executed, the script produces a report grouped by author, making it easy to find out who owns stale branches:
Author: Alice
- repo1 -> feature/login-page
- repo2 -> hotfix/payment-fix
Author: Bob
- repo3 -> refactor/api-updates
- repo1 -> test/legacy-integration
Author: Unknown
- repo4 -> bugfix/session-timeout
🔥 Why This Matters
- Easier Maintenance – Identify branches that can be deleted or merged.
- Better Collaboration – Reach out to authors to confirm branch status.
- Improved Performance – Reducing unnecessary branches speeds up repository operations.
You can extend this script further to:
✅ Filter branches by last commit date
✅ Automatically delete stale branches (with gh api -X DELETE)
✅ Generate a GitHub issue or PR listing stale branches
Would love to hear how you customize this for your workflow! 🚀