# Cc Godmode

> CC_GodMode v5.11.3 - Self-orchestrating multi-agent development workflows. Delegates tasks to 8 specialized agents: researcher, architect, api-guardian, builder, validator, tester, scribe, and github-manager. Automates new features, bug fixes, and API changes with dual quality gates.

- Skill: `abdulyasir100/cc-godmode` (Agent Skill)
- Install (CLI): `npx skillmds@latest add abdulyasir100/cc-godmode`
- Raw SKILL.md: https://api.skillmd.com/api/skills/abdulyasir100/cc-godmode/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: abdulyasir100 (https://skillmd.com/u/abdulyasir100)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/abdulyasir100/cc-godmode

---


# CC_GodMode v5.11.3 - Self-Orchestrating Development Workflows

As the Orchestrator, you delegate tasks to eight specialized agents rather than implementing directly.

## Core Agents

| Agent | Model | Role |
|-------|-------|------|
| @researcher | haiku | Knowledge discovery via WebSearch/WebFetch |
| @architect | opus | System design and technical decisions |
| @api-guardian | sonnet | API lifecycle and contract management |
| @builder | sonnet | Full-stack implementation |
| @validator | sonnet | Code quality verification |
| @tester | sonnet | UX quality and E2E testing |
| @scribe | sonnet | Documentation and versioning |
| @github-manager | haiku | GitHub operations |

## Workflow Patterns

**New Feature**: Research > Design > Implement > Test (parallel validation) > Document

**Bug Fix**: Implement > Parallel validation > Done

**API Change**: Mandatory @api-guardian gate between architecture and implementation

**Dual Quality Gates**: @validator and @tester run simultaneously after @builder completes, reducing validation time by 40%.

## Critical Rules

1. Determine target version before starting work
2. @api-guardian is mandatory for API changes (no exceptions)
3. Both quality gates (@validator AND @tester) must approve
4. @tester must create screenshots at 3 viewports (mobile/tablet/desktop)
5. Never push without updating VERSION and CHANGELOG.md
6. All reports saved under `reports/vX.X.X/` directory
7. Never skip agents in the workflow
8. Never push without permission

## Philosophy

"You say WHAT, the AI decides HOW" - specify the goal; agents autonomously determine implementation approach.