README
CC_GodMode
"What happens when an AI system is used to improve itself?"
You're looking at the answer.
The Story
It started simple: One developer, mass sleep deprivation, and a vision.
Phase 1: Manual labor. Researching best practices. Reading docs. Testing prompts. Failing. Iterating. Building agent after agent. Workflow after workflow. Week after week.
Phase 2: The system works. 8 specialized AI agents orchestrating themselves. Features get built. Bugs get fixed. Documentation writes itself. "This is pretty good," I thought.
Phase 3: January 6th, 2026. A thought: "What if I use the system... to improve the system?"
I gave it one prompt. The orchestrator delegated to the research team. Analyzed its own architecture. Found inefficiencies. Proposed improvements. Implemented them. Validated itself. Documented the changes.
The loop closed.
Phase 4: You're reading this README. An AI wrote parts of it. An AI will improve it. The experiment continues.
What Is This?
CC_GodMode transforms Claude Code into a self-orchestrating development team.
You say WHAT. The AI figures out HOW.
You: "I need user authentication with JWT"
Orchestrator:
→ Analyzes request
→ Determines version (5.5.0)
→ Creates report folder
→ Delegates to @architect for design
→ Delegates to @api-guardian for API impact
→ Delegates to @builder for implementation
→ @validator checks code quality
→ @tester checks UX quality
→ @scribe documents everything
→ @github-manager creates PR
You: *drinks coffee*
The difference?
| Without CC_GodMode | With CC_GodMode |
|---|---|
| You: "Design the feature" | You: "Build Feature X" |
| You: "Now implement it" | ☕ |
| You: "Check the types" | ☕ |
| You: "Update the consumers" | ☕ |
| You: "Write the docs" | ☕ |
| You: "Did I forget something?" | AI: "Done. Here's the report." |
The Agents
8 specialists. Each with their own expertise. Each knowing exactly what they do—and what they don't.
| Agent | Role | Specialty |
|---|---|---|
@researcher | Knowledge Discovery | Web research, documentation lookup, technology evaluation (NEW v5.10.0) |
@architect | System Architect | High-level design, module structure, tech decisions |
@api-guardian | API Lifecycle Expert | Breaking changes, consumer impact, contract validation |
@builder | Senior Developer | Implementation, following @architect's specifications |
@validator | Code Quality Gate | TypeScript, unit tests, security, consumer verification |
@tester | UX Quality Gate | E2E tests, visual regression, accessibility, performance (Enhanced v5.10.0) |
@scribe | Technical Writer | Documentation, changelog, version management |
@github-manager | GitHub Manager | Issues, PRs, releases, CI/CD orchestration |
Dual Quality Gates:
@builder completes
│
┌───────────────┴───────────────┐
▼ ▼
┌─────────────┐ ┌─────────────┐
│ @validator │ │ @tester │
│ Code Quality│ │ UX Quality │
├─────────────┤ ├─────────────┤
│ ✓ TypeScript│ │ ✓ E2E Tests │
│ ✓ Unit Tests│ │ ✓ Visuals │
│ ✓ Security │ │ ✓ A11y │
│ ✓ Consumers │ │ ✓ Perf │
└──────┬──────┘ └──────┬──────┘
│ │
└───────────────┬───────────────┘
▼
Both gates passed?
→ Continue to @scribe
The Architecture (v6.3)
v6.0 introduced modular architecture. v6.1 added Skills. v6.2 added worktree isolation. v6.3 added Plugin packaging.
~/.claude/ ← RUNTIME (What Claude loads)
├── agents/ ← 8 agents, globally available
│ ├── researcher.md
│ ├── architect.md
│ ├── api-guardian.md
│ ├── builder.md
│ ├── validator.md
│ ├── tester.md
│ ├── scribe.md
│ └── github-manager.md
├── scripts/ ← Hook scripts
│ ├── session-start.js
│ ├── validate-agent-output.js
│ └── check-api-impact.js
└── settings.json ← Hooks (SessionStart, PostToolUse,
SubagentStop, TaskCompleted)