GitHub Issue-to-Pull-Request Agent

The problem

GitHub issues are often implementation-ready. Developers waste time interpreting ambiguous requirements, rediscovering repository patterns, and writing repetitive PR descriptions instead of focusing on code.

What it does

• Extracts acceptance criteria and behavior models from GitHub issues and comments.
• Performs repository reconnaissance to identify impacted files and local coding conventions.
• Generates targeted code changes and automated tests that follow existing architecture.
• Performs multi-layered verification including security, privacy, and accessibility reviews.
• Produces complete pull request packages with titles, descriptions, and reviewer guidance.

Frameworks & tools

Works with GitHub, Cursor, Claude Code, and Copilot. Supports JavaScript, TypeScript, Python, Go, Rust, Java, Swift, Ruby, SQL, Terraform, and Docker.

Why this beats prompting it yourself

Standard prompts often result in scope creep, unrelated refactoring, or tests that validate implementation instead of behavior. This skill enforces a disciplined sequence from requirement modeling to evidence-based verification, ensuring minimal diffs and review-ready outputs.

Use cases

• Convert vague bug reports into reproducible test cases and verified fixes.
• Decompose large, complex issues into manageable, sequenced pull requests.
• Generate comprehensive API or UI feature implementations with migration notes.
• Audit existing diffs against original issue requirements for compliance.

Known limitations

Cannot independently guarantee deployment safety without repository access and executed CI evidence. High-impact architectural or security decisions require manual specialist approval.