project-creator

Project Creator for CCGG Business Operations

Quick Start

What this skill does: Automatically creates new CCGG Business Operations projects with all required mechanisms pre-implemented (PARENT SYSTEM INTEGRATION, Active Projects Index, operations logging, etc.).

Use when: Creating new incubator project, adding sub-project to CCGG, setting up project structure

Prevents: Forgetting critical mechanisms like PARENT SYSTEM INTEGRATION (the problem that triggered this skill's creation)

When to Use This Skill

Trigger Phrases:

• "Create new project in CCGG Business Operations"
• "Initialize new incubator project"
• "Set up project structure for [project-name]"
• "Add new sub-project to CCGG"

Project Types Supported:

1. Simple Projects (majority): Research, single-purpose tools, straightforward deliverables
2. Complex Projects (strategic): Multi-project coordination, dependencies, integration requirements

Project Creation Workflow

Step 1: Gather Project Information

Ask Daron These Questions (in conversational order):

1. "What's the project name?"
2. Format: hyphen-case (e.g., member-retention-sequences)
3. Max 40 characters
4. Will be used for folder name and project ID

1. "What's the project purpose?" (1-2 sentences)
2. Clear problem statement
3. Expected outcome
4. Example: "Design email sequences to re-engage churned members and improve retention rate"

1. Analyze project description for multi-phase indicators (SUGGESTION, not question):

Claude analyzes project purpose and deliverables for signals:

• Keywords: "validate", "test", "feedback", "iterate", "production rollout", "pilot"
• Complex deliverables (3+ major components requiring testing)
• Integration with existing systems (needs testing phase)
• User mentions "experiment", "trial", or "phase"
• Project type: infrastructure, framework, new system (vs simple content creation)

If multi-phase indicators detected, SUGGEST phased approach: ``` "Based on your project description, I recommend a multi-phase approach:

Phase 1: [Research & Planning / Setup & Test / MVP] Phase 2: [Build & Test / Validation & Refinement / Production Rollout] Phase 3: [Deploy & Scale / Institutionalize / Maintenance]

This allows testing and feedback before full rollout. Would you like to use this phased approach?" ```

User can respond:

• "Yes, use phases" → Generate PHASE_TRACKER.md with suggested phases
• "No, single phase" → Skip Phase Tracker
• "Let me customize phases" → Ask for phase names/durations

If NO multi-phase indicators, skip suggestion and Phase Tracker

1. "Will this project coordinate with or depend on other projects?"
2. If YES → Complex project (needs Coordination Hub)
3. If NO → Simple project (standard structure only)

1. ALWAYS: Detect and Capture Dependencies (NEW - Forcing Function)

Claude PROACTIVELY suggests dependencies based on project description:

``` "Let me help identify dependencies for this project:

BLOCKING DEPENDENCIES (must complete before starting this):

• What existing work must finish before you can start?
• What deliverables from other projects do you need?

DOWNSTREAM DEPENDENCIES (projects waiting for this):

• What other projects are waiting for this one?
• What will this project enable or unblock?

RELATED PARALLEL (connected but not blocking):

• What other projects share themes/tools with this?

[Based on your project description, I see potential dependencies:]

• [Suggest upstream based on keywords/deliverables mentioned]
• [Suggest downstream based on Active Projects needing this work]

Should I add these to the dependency metadata?" ```

Capture in YAML format: ``yaml dependencies: blocks: ["project-id-1", "project-id-2"] # Must complete first blocked_by: ["project-id-3"] # Waiting for this related_parallel: ["project-id-4"] # Connected, not blocking ``

Even if "none", still create empty fields (forces conscious decision): ``yaml dependencies: blocks: [] # No upstream dependencies blocked_by: [] # No downstream dependencies related_parallel: [] # No related work ``

1. If Complex: "Which projects does this coordinate with?"
2. Upstream dependencies (what this project needs)
3. Downstream dependencies (what other projects need from this)
4. Example: "Depends on hormozi-money-models (frameworks), feeds into retention-reengagement (implementation)"
5. NOTE: This is for Coordination Hub documentation (prose), Step 5 captures YAML metadata

1. "What are the key deliverables?" (3-5 items)
2. Specific outputs this project will produce
3. Example: "Email sequence templates, DM scripts, retention playbook"

1. "Which avatars does this serve?" (optional, for strategic alignment)
2. From Target_Avatars_Complete_Profiles.md
3. If "all avatars" → note that
4. If specific → list them

Step 2: Determine Project Complexity

Based on Question 3 answer:

SIMPLE PROJECT (if NO coordination):

• Standard folder structure
• CLAUDE.md with PARENT SYSTEM INTEGRATION
• README.md
• Active Projects Index
• operations_log entry

COMPLEX PROJECT (if YES coordination):

• All simple project components PLUS:
• Coordination Hub/ folder
• PROJECT_DEPENDENCIES.md
• INTEGRATION_CHECKLIST.md
• OUTPUT_LIBRARY.md
• Enhanced CLAUDE.md (cross-project knowledge access patterns)
• Integration sections in related projects' CLAUDE.md files

Step 3: Create Project Structure

Location: Active Projects/_Incubator/[project-name]/

Manual Creation (automated script planned for future):

1. Create folder: mkdir "Active Projects/_Incubator/[project-name]"
2. Use Write tool to create CLAUDE.md from templates/CLAUDE_SIMPLE.md OR templates/CLAUDE_COMPLEX.md
3. If multi-phase project: Use Write tool to create PHASE_TRACKER.md from templates/PHASE_TRACKER_TEMPLATE.md
4. Fill all placeholders (marked with {{VARIABLE}})

This creates:

[project-name]/
├── CLAUDE.md                    # Project guidance (PARENT SYSTEM INTEGRATION included)
├── README.md                    # Quick start + overview
├── [folders based on project type]
└── Coordination Hub/            # If complex project
    ├── PROJECT_DEPENDENCIES.md
    ├── INTEGRATION_CHECKLIST.md
    └── OUTPUT_LIBRARY.md

Note: Full automation script (create_project.sh) is planned but not yet implemented. Current workflow uses templates + manual variable replacement.

Step 4: Generate CLAUDE.md Content

Use Template:

• Simple: templates/CLAUDE_SIMPLE.md
• Complex: templates/CLAUDE_COMPLEX.md

Required Sections (ALL projects):

1. PROJECT IDENTITY (name, type, status, created date, owner)
2. PROJECT MISSION (purpose, core focus, expected outcomes)
3. SCOPE & BOUNDARIES (in scope, out of scope)
4. PARENT SYSTEM INTEGRATION ⚠️ CRITICAL
5. Project Memory Index Sync (path, when/how to update)
6. Operations Logging (format, actions to auto-log, examples)
7. Strategic Alignment Validation (OOBG check, UV check, Avatar targeting)
8. Cross-Project Intelligence (search related projects)
9. PROJECT STRUCTURE (folder organization, key files)
10. EXPECTED DELIVERABLES (phased if applicable)
11. SUCCESS CRITERIA (how to measure completion)

Additional Sections (Complex projects):

1. AUTONOMOUS KNOWLEDGE ACCESS (cross-project references, access commands)
2. COORDINATION WITH OTHER PROJECTS (dependency chain, integration points)

Populate Variables:

• Replace {{PROJECT_NAME}} with actual name
• Replace {{PROJECT_PURPOSE}} with purpose from Step 1
• Replace {{DELIVERABLES}} with list from Step 1
• Replace {{UPSTREAM_DEPS}} and {{DOWNSTREAM_DEPS}} with projects from Step 1 (if complex)
• Repl