Effective Claude workflows - /solo for 90% of tasks, /team for complex features. Based on official Anthropic patterns.
/plugin marketplace add ashchupliak/dream-team
/plugin install dream-team@ashchupliak/dream-team
Build the project with quality checks
Run Codex CLI for code analysis or generation tasks
Deploy application to specified environment (staging/prod)
Discover and analyze repository to generate context files for the team
Emergency hotfix workflow for production issues
Investigate an issue using multiple specialized agents in parallel
Start a new feature implementation with full team coordination
Run parallel code review with multiple specialists (code-reviewer + security-tester)
Fast 3-agent workflow for quick fixes and small changes (analyst → developer → qa)
Perform code review on recent changes or specified PR
Run comprehensive security audit on the codebase
Run tests with optional scope (unit/integration/e2e/all)
Effective Claude workflow - builds context, asks when needed, works incrementally
Team-A (Full) - 5-agent team with health checks and strict protocol
Team-B (Standard) - 5-agent team with simplified workflow
Team-C (Fast) - 3-agent lightweight team for quick iterations
Intelligent Engineering Manager - 7-phase feature development with parallel agents and user checkpoints
Automated code review for pull requests using multiple specialized agents with confidence-based scoring
Claude Agent SDK Development Plugin
Implementation of the Ralph Wiggum technique - continuous self-referential AI loops for interactive iterative development. Run Claude in a while-true loop with the same prompt until task completion.
Comprehensive PR review agents specializing in comments, tests, error handling, type design, code quality, and code simplification