Claude Code for Continuous Integration
Your CI pipeline uses a Claude Code agent to triage incoming pull requests: it labels PRs by type (bug fix, feature, refactor, docs), assigns reviewers, and runs targeted test suites based on label. Three weeks after deployment, the team reports that the agent frequently mis-labels PRs that combine a bug fix with a refactor, causing the wrong test suites to run and reviewers to be misassigned. Logs show the agent receives the full PR diff and description correctly. What is the most effective architectural change to address this?