Exporting to GitHub

feedback:github turns a feedback item — or every item in a cluster — into a GitHub issue. Run it when you're ready to move actionable feedback into your normal dev workflow.

Setup

1. Set your environment variables

The command reads credentials from the environment. Add these to your .env file (or however you manage secrets in your deployment):

GITHUB_TOKEN=ghp_yourPersonalAccessToken
GITHUB_OWNER=your-org-or-username
GITHUB_REPO=your-repo-name

Your token needs the repo scope (or public_repo for public repositories).

Never hard-code credentials

Don't put your token in app/Config/Betta.php. The config properties ($githubToken, $githubOwner, $githubRepo) exist as a fallback for unusual setups, but .env is the right place for credentials.

2. Create labels in your GitHub repo

The command applies labels to each issue — it won't create them. Before you start exporting, make sure these labels exist in your repo:

Label	When it's applied
bug	Feedback category is bug
ux	Feedback category is ux
feature	Feedback category is feature
other	Feedback category is other
low, medium, high, critical	Cluster priority (when exporting a whole cluster)

You only need the labels that match your actual data — missing labels are silently skipped, not an error.

Basic usage

Export a single feedback item by its ID:

php spark feedback:github 42

That creates one GitHub issue and writes the issue URL back to the betta_feedback row. You'll see the link printed in the terminal:

Feedback #42 → https://github.com/your-org/your-repo/issues/123

Cluster mode

Export every item in a cluster at once with --cluster:

php spark feedback:github 7 --cluster

This creates one GitHub issue per feedback item in cluster 7 — not one issue for the cluster. Each item becomes its own trackable ticket so you can close them independently as you fix things.

Feedback #14 → https://github.com/your-org/your-repo/issues/124
Feedback #27 → https://github.com/your-org/your-repo/issues/125
Feedback #31 → https://github.com/your-org/your-repo/issues/126

--dry-run

Not sure what you're about to create? Use --dry-run to print a preview without touching the GitHub API:

php spark feedback:github 7 --cluster --dry-run

Output looks like:

Would create issue for feedback #14:
Title:  [Feedback] bug: Login button doesn't respond on mobile Safari
Body:
Login button doesn't respond on mobile Safari — tapping it does nothing.

---
**Category:** bug
**Email:** user@example.com
**URL:** https://app.example.com/login
**Submitted:** 2026-05-12 09:14:33
Labels: bug, high

Dry-run works without credentials set — useful for previewing on a CI machine before a real run.

How issues are formatted

Every issue has the same structure:

Title: [Feedback] {category}: {first 80 characters of the message}

Body:

{full message}

---
**Category:** bug
**Email:** user@example.com        (omitted if blank)
**URL:** https://app.example.com   (omitted if blank)
**Sentiment:** 1                   (omitted if not recorded)
**Submitted:** 2026-05-12 09:14:33

Labels: the category value (bug, ux, feature, or other), plus the cluster's priority label if you exported via --cluster.

Re-running safely

If a feedback item already has a github_issue_url, the command skips it and shows you the existing link:

Feedback #14 already exported: https://github.com/your-org/your-repo/issues/124

This means you can safely re-run the command on a cluster after adding new items — previously exported items are untouched, and only the new ones get issues created.

Next steps

• Managing Clusters — set priorities on clusters before exporting so the right labels land on your GitHub issues
• AI Clustering — use AI to group ungrouped feedback before exporting