Multi-agent plugin that analyzes codebases to find ways to achieve the same functionality with less code. Uses haiku for fast review tasks and sonnet for deep analysis. Always starts in plan mode, never changes UI/UX or language.
Autonomous AI sandbox that builds whatever it wants in complete isolation. Runs continuously in Docker with creative freedom while staying safely contained.
Multi-agent Claude Code system for disciplined Salesforce user story generation with Slack-based approval workflows, architectural review, and quality gates
Autonomous engineering organization with 26+ AI agents. Build features, review code, and track costs automatically.