fix

Automated code fixes for GitHub issues require implementing minimal changes identified by root-cause analysis and adding regression tests. This process ensures fixes do not introduce new issues by running targeted unit tests, typecheck, and i18n checks before finalizing changes.

Root-cause analysis in GitHub issue management identifies the minimal code change needed to resolve a problem. It precedes the automated fix implementation step, providing the exact code modifications required to address the underlying issue.

Automating code fixes requires write access to the repository. This permission is necessary for claiming GitHub issues to prevent concurrent automation, implementing code changes, and adding regression tests directly to the project.

Preventing concurrent automation involves claiming the GitHub issue at the start of the fix process. This locks the issue so other automated systems do not duplicate work, and it is released later in the open-pr step.

Validation steps after implementing automated code fixes include running targeted unit tests, typecheck, i18n checks, and project-specific validation steps. The changes also undergo self-review through code-review and backward-compatibility policy checks.

Automated code fixing works with project-specific validation rules by running targeted unit tests, typecheck, i18n checks, and other validation steps defined in the project. It requires knowledge of the project's AGENTS.md rules to ensure compliance.