New Step 6: Configure Ralph CLI Autonomous Looping (Advanced - Optional)
- Install Ralph CLI: npm install -g @iannuttall/ralph
- Create ralph-auto-trigger.sh hook script
- Configure hooks.json for auto-triggering
- Control modes: always/agents/off (default: agents)
- State persistence in .claude/ralph-loop.local.md
- Cancel methods: /cancel-ralph, delete state file, set mode=off
Key Features:
- Autonomous looping until task completion
- Multi-iteration workflows
- State management across loops
- Integrates with PROACTIVELY agents in HYBRID mode
Updated Step 7: Verify Complete Installation
- Added Ralph CLI verification
- Added Ralph hook verification
- Marked Ralph as optional (⚠️)
Updated Summary:
- Changed 'Two Auto-Triggering Systems' to 'Three'
- Added Ralph CLI as third system
- Clarified PROACTIVELY vs Ralph CLI vs Hooks
HYBRID Approach:
- Users can choose patterns-only (simple)
- Or Ralph CLI (advanced autonomous)
- Or combine both for maximum power
- Default: PROACTIVELY patterns built-in
- Optional: Ralph CLI for complex workflows
Changes:
- Remove Ralph from list of installed components (it was misleading)
- Change component count from 7 to 6
- Add Agent Architecture Patterns section explaining:
- PROACTIVELY coordinators implement proven coordination patterns
- Patterns inspired by frameworks like Ralph
- Supervisor-agent delegation and shared context management
- Clarify Ralph is a conceptual reference, not an installed tool
This prevents confusion - Ralph CLI is not installed, only its
coordination patterns influenced the agent design.
MASTER-PROMPT.md restructure:
- Master prompt now in ONE continuous code block (bash blocks to plain text
- Added clear labels for code sections within the prompt
- Real-life examples now in proper markdown outside code block
- Configuration examples in separate code blocks after main prompt
- Quick reference section properly formatted
- All content preserved, just restructured for proper rendering
The master prompt can now be copied as a single block
without formatting breaking mid-copy.
- Update PROACTIVELY agents from 5 to 7 (add studio-producer, project-shipper)
- Add clear architecture overview: 7 coordinators + 31 specialists
- Explain two pathways: automatic vs direct control
- Add real workflow example showing coordinator orchestration
- Clarify how coordinators call specialists automatically
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
- Replace HTML div tags with proper markdown formatting
- Use box-drawing characters for visual separation
- Add clear section headers with emoji indicators
- Use code blocks with language tags (diff for ❌ examples)
- Add proper spacing and horizontal rules
- Ensure GitHub renders all formatting correctly
- Make examples section clearly distinct from main content
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
- Create prominent boxed section with gradient background
- Add clear warning box explaining these are illustrative examples only
- Use color-coded boxes: yellow (❌ WITHOUT) vs green (✅ WITH)
- Add proper visual separation from main installation content
- Include border styling and padding for clear visual hierarchy
- Make it immediately clear this is a separate reference section
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
- Create dedicated 'How Auto-Triggering Works: Practical Examples' section
- Add clear NOTE explaining these are illustrative examples
- Consolidate 5 examples showing WITHOUT vs WITH scenarios
- Add visual separation with headers and dividers
- Include summary table comparing without vs with auto-triggering
- Remove duplicate examples scattered throughout document
- Make it immediately clear these are workflow examples
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
- Add llm-tldr (95% token reduction) to all installation methods
- Update MASTER-PROMPT.md with all 6 sources and real-life examples
- Update README.md with comprehensive source guide
- Update automation script with TLDR installation and MCP config
- Update INTEGRATION-GUIDE.md to clarify Z.AI MCP tools work with GLM
- Add MCP explanation and link to modelcontextprotocol.io
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
BREAKING CHANGES to workflow:
- Add 'BEFORE YOU BEGIN' section to MASTER-PROMPT.md
- GLM users MUST configure GLM FIRST before using Claude Code
- Add clear GLM pre-configuration steps at top of file
- Add clear copy boundaries (✂️ COPY FROM HERE / ✂️ COPY ENDS HERE)
- Add visual separator lines for copy-paste section
README.md updates:
- Add prominent GLM warning at top of Installation Options
- Add quick GLM setup commands
- Clarify Master Prompt workflow with step-by-step
- Emphasize GLM pre-configuration requirement
This fixes the issue where GLM users couldn't use Claude Code
to paste the master prompt without first configuring GLM.
- Add Claude Code installation prerequisite (npm install)
- Update GLM configuration with official model mappings (glm-4.5-air, glm-4.7)
- Add version verification (recommended 2.0.14+)
- Include troubleshooting steps from Z.AI docs
- Add /status command verification
- Reference official Z.AI documentation: https://docs.z.ai/devpack/tool/claude
- Update README with proper GLM model configuration
- Users can now copy-paste a master prompt into Claude Code
- Includes all source repository references
- Step-by-step guided installation
- Alternative to running bash scripts
- Updated README with 3 installation options
·