More than 15,000 organizations rely on Codacy to review their pull requests against pre-defined quality, security, and AI coding standards. That feedback has always lived inside the PR view: quality and security issues found, drops in test coverage, intent gaps, and fix suggestions.
That worked well in the pre-AI era. Now that coding agents have multiplied how much code gets written, code review has become the bottleneck for teams adopting agentic workflows. The scans still run automatically, but triaging the findings takes more human effort to clear the merge: fixing critical issues, opening Jira tickets to snooze the important-but-not-urgent fixes, and tagging false positives with an ignore-reason.
Today we're shipping Codacy Skills, a new way to let coding agents like Claude handle that gruntwork to unblock pull requests faster, configure Codacy rules and settings, and even perform the code analysis locally pre-commit.