Automating Feature-Flag Cleanup at Scale with a Multi-Agent LLM System

DoorDash built a multi-agent LLM system to automate the removal of stale feature flags across thousands of repositories, addressing a significant technical debt backlog. The system combines an orchestrator agent using live platform metadata to determine target values with parallel removal agents running in isolated Git worktrees. Tested on fifty recent stale flags, it successfully produced merge-ready pull requests for ninety percent of them at an average cost of $4.79 and 13.8 minutes per cleanup. Open article →