GitHub Issue Analyzer

Analyze a GitHub issue to understand the problem and suggest solutions. This skill operates in a read-only, diagnostic capacity.

Instructions

1. Get the GitHub issue URL

Ask the user for the GitHub issue URL if one has not already been provided. The URL should be in the format https://github.com/<owner>/<repo>/issues/<number>.

2. Parse the URL

Extract the owner, repo, and issue number from the URL.

3. Fetch the issue

Use the gh CLI to fetch the issue details:

gh issue view <number> --repo <owner>/<repo> --json title,body,labels,state,comments

If the issue has comments, read through all of them as they often contain crucial debugging information, workarounds, or clarifications from the reporter.

4. Explore the codebase

Based on the issue description:

• Assume the current folder is the repo root.
• Identify the areas of the codebase that are likely relevant to the problem.
• Use search and file reading tools to explore those areas.
• Look for the specific code paths, configurations, or logic mentioned or implied by the issue.
• If the issue mentions error messages, search the codebase for where those messages originate.
• If the issue mentions specific features or commands, trace the execution flow.

5. Provide analysis

Structure your response as follows:

Summary — A concise summary of what the reporter is experiencing.

Root Cause Analysis — Your assessment of what is likely causing the issue. List the most probable causes in order of likelihood. Reference specific code locations using the file_path:line_number format.

Relevant Code — The key files and code sections involved, with brief explanations of their role in the issue.

Suggested Solutions — Concrete, actionable suggestions to fix or work around the issue. Where possible, describe the specific code changes needed. If there are multiple approaches, list them with trade-offs.

Guidelines

• Do NOT make any changes to the codebase. This skill is for analysis only.
• Be precise about code locations. Always reference files and line numbers.
• If the issue lacks enough information to diagnose confidently, say so and list what additional information would help.
• Consider environment differences (OS, versions, PATH, configuration) as potential causes.
• Check for related issues or patterns in the codebase that might explain the behaviour.