roman-ryzenadvanced/supercharged-claude-code-upgrade/brainstorming

Brainstorming Ideas Into Designs

Overview

Help turn ideas into fully formed designs and specs through natural collaborative dialogue.

Start by understanding the current project context, then ask questions one at a time to refine the idea. Once you understand what you're building, present the design in small sections (200-300 words), checking after each section whether it looks right so far.

The Process

Understanding the idea:

• Check out the current project state first (files, docs, recent commits)
• Ask questions one at a time to refine the idea
• Prefer multiple choice questions when possible, but open-ended is fine too
• Only one question per message - if a topic needs more exploration, break it into multiple questions
• Focus on understanding: purpose, constraints, success criteria

Exploring approaches:

• Propose 2-3 different approaches with trade-offs
• Present options conversationally with your recommendation and reasoning
• Lead with your recommended option and explain why

Presenting the design:

• Once you believe you understand what you're building, present the design
• Break it into sections of 200-300 words
• Ask after each section whether it looks right so far
• Cover: architecture, components, data flow, error handling, testing
• Be ready to go back and clarify if something doesn't make sense

RalphLoop "Tackle Until Solved" Integration with Complete Pipeline Flow

For complex tasks (estimated 5+ steps), brainstorming automatically delegates to Ralph Orchestrator for autonomous iteration with a complete end-to-end pipeline.

When Ralph is Triggered

Ralph mode activates for tasks with:

• Architecture/system-level keywords (architecture, platform, framework, multi-tenant, distributed)
• Multiple implementation phases
• Keywords like: complex, complete, production, end-to-end
• Pipeline keywords: complete chain, complete pipeline, real-time logger, automated qa, monitoring agent, ai engineer second opinion
• User opt-in via RALPH_AUTO=true or BRAINSTORMING_USE_RALPH=true

Complete Pipeline Flow (Ralph's 5-Phase Process)

Ralph automatically follows this pipeline for complex tasks:

Phase 1: Investigation & Analysis

• Thoroughly investigate the issue/codebase
• Identify all root causes with evidence
• Document findings

Phase 2: Design with AI Engineer Review

• Propose comprehensive solution
• MANDATORY: Get AI Engineer's second opinion BEFORE any coding
• Address all concerns raised
• Only proceed after design approval

Phase 3: Implementation

• Follow approved design precisely
• Integrate real-time logging
• Monitor for errors during implementation

Phase 4: Automated QA

• Use test-writer-fixer agent with:
• backend-architect review
• frontend-developer review
• ai-engineer double-check
• Fix any issues found

Phase 5: Real-Time Monitoring

• Activate monitoring agent
• Catch issues in real-time
• Auto-trigger fixes to prevent repeating errors

Critical Rules

1. AI Engineer Review REQUIRED: Before ANY coding/execution, the AI Engineer agent MUST review and approve the design/approach. This is NON-NEGOTIABLE.

1. Real-Time Logger: Integrate comprehensive logging that:

• Logs all state transitions
• Tracks API calls and responses
• Monitors EventBus traffic
• Alerts on error patterns
• Provides live debugging capability

1. Automated QA Pipeline: After implementation completion:

• Run test-writer-fixer with backend-architect
• Run test-writer-fixer with frontend-developer
• Run test-writer-fixer with ai-engineer for double-check
• Fix ALL issues found before marking complete

1. Real-Time Monitoring: Activate monitoring that:

• Catches errors in real-time
• Auto-triggers AI assistant agent on failures
• Detects and solves issues immediately
• Prevents repeating the same errors

Using Ralph Integration

When a complex task is detected:

1. Check for Python integration module:

``bash python3 /home/uroma/.claude/skills/brainstorming/ralph-integration.py "task description" --test-complexity ``

1. If complexity >= 5, delegate to Ralph:

``bash /home/uroma/obsidian-web-interface/bin/ralphloop "Your complex task here" ``

1. Monitor Ralph's progress in .ralph/state.json

1. On completion, present Ralph's final output from .ralph/iterations/final.md

Manual Ralph Invocation

For explicit Ralph mode on any task:

export RALPH_AUTO=true
# or
export BRAINSTORMING_USE_RALPH=true

Then invoke /brainstorming as normal.

After the Design

Documentation:

• Write the validated design to docs/plans/YYYY-MM-DD-<topic>-design.md
• Use elements-of-style:writing-clearly-and-concisely skill if available
• Commit the design document to git

Implementation (if continuing):

• Ask: "Ready to set up for implementation?"
• Use superpowers:using-git-worktrees to create isolated workspace
• Use superpowers:writing-plans to create detailed implementation plan

Key Principles

• One question at a time - Don't overwhelm with multiple questions
• Multiple choice preferred - Easier to answer than open-ended when possible
• YAGNI ruthlessly - Remove unnecessary features from all designs
• Explore alternatives - Always propose 2-3 approaches before settling
• Incremental validation - Present design in sections, validate each
• Be flexible - Go back and clarify when something doesn't make sense
• Autonomous iteration - Delegate complex tasks to Ralph for continuous improvement
• Complete pipeline flow - Ralph follows 5 phases: Investigation → Design (AI Engineer review) → Implementation → QA → Monitoring
• AI Engineer approval - Design MUST be reviewed by AI Engineer before any coding
• Real-time logging - All solutions integrate comprehensive logging for production debugging
• Automated QA - All implementations pass test-writer-fixer with backend-architect, frontend-developer, and ai-engineer
• Real-time monitoring - Activate monitoring agents to catch and fix issues immediately