Skip to content

GitHub Issue Auto-Labeller

Example prompt: "When a new GitHub issue is opened, classify it as bug, feature, or docs, add the right label, assign it to the correct team, and post a comment acknowledging receipt."

How to automate GitHub issue labelling with GloriaMundo

The Problem

Open-source projects and internal repositories alike suffer from unlabelled, unassigned issues that pile up in the backlog. A new issue arrives with a vague title and a wall of text, and someone — usually a senior engineer or maintainer — has to read it, decide whether it is a bug report, a feature request, or a documentation gap, apply the right labels, and route it to the correct team. For repositories receiving 10-20 new issues a day, this triage work eats into time that should be spent writing code. Issues that are not labelled quickly tend to be forgotten entirely.

How GloriaMundo Solves It

We build a workflow that triggers whenever a new issue is opened on a GitHub repository. An integration step fetches the issue's title, body, and any linked context. An LLM step reads the content and classifies it into a category — bug, feature, docs, or question — based on your project's labelling conventions. A conditional step checks the classification and determines which team or individual should be assigned. Integration steps then apply the appropriate labels, set the assignee, and post a comment on the issue acknowledging receipt and letting the reporter know which team is looking at it. Glass Box preview shows you the proposed label, assignee, and comment text before any changes are made to the issue.

Example Workflow Steps

  1. Trigger (webhook): Fires when a new issue is opened on the target GitHub repository.
  2. Step 1 (integration): Fetch the issue title, body, author, and any linked pull requests or related issues from GitHub.
  3. Step 2 (LLM): Classify the issue as bug, feature, docs, or question based on its content, and suggest a priority level.
  4. Step 3 (conditional): Route the issue to the correct team based on the classification — bugs to the platform team, features to product, docs to the docs maintainers.
  5. Step 4 (integration): Apply the classification label and priority label to the issue on GitHub.
  6. Step 5 (integration): Assign the issue to the designated team lead.
  7. Step 6 (integration): Post a comment on the issue acknowledging receipt and confirming which team will review it.

Integrations Used

  • GitHub — source of the new issue and target for labels, assignees, and acknowledgement comments
  • Slack — optional notification to the assigned team's channel when a high-priority issue is triaged

Who This Is For

Open-source maintainers, engineering managers, and DevOps teams who manage repositories with a steady flow of incoming issues and want consistent, fast triage without relying on a single person to read every new report.

Time & Cost Saved

Manual issue triage takes roughly 2-5 minutes per issue. For a repository receiving 15 new issues a day, that is 30-75 minutes of daily triage work. This workflow handles the classification and labelling in seconds, reducing the human effort to a quick check of the Glass Box preview for edge cases. Over a month, a busy repository could reclaim 10-25 hours of maintainer time. The workflow uses integration, LLM, and conditional steps, costing a few credits per issue.