claude-fulcrum

Cross-platform AI dev stack — 25 agents, 108 skills across Claude Code, Copilot, Cursor, Codex & OpenCode with shared orchestration via claude-flow

25 Agents · 108 Skills · 57 Commands · 9 Language Rule Sets · Multi-Agent Orchestration

Quick Start · How It Differs From ECC · Cross-Platform · Features · Guides · Architecture

---

Built after integrating ECC, claude-flow, and GitHub Copilot into a single shared-memory stack. This is the reference implementation for running all 5 major AI coding tools together — with one orchestration layer, one vector memory, and one set of agents.

---

🚀 Quick Start

git clone https://github.com/ORION2809/claude-Fulcrum.git
cd claude-fulcrum
npm install
./install.sh typescript   # replace with your stack (python, golang, kotlin, etc.)

Windows: .\install.ps1 typescript

npm test   # verify everything works

Other installation methods

Claude Code Plugin

/plugin install claude-fulcrum

npx (Quick Try)

npx claude-fulcrum typescript

---

⚖️ How This Differs From ECC

If you've used Everything Claude Code, here's what claude-fulcrum adds:

	ECC	claude-fulcrum
Platforms	4 (Claude Code, Codex, Cursor, OpenCode)	5 (+ full GitHub Copilot)
Language rule sets	5	9 (+ Kotlin, Rust, C++, Perl)
Agents	21	25
Multi-agent orchestration	Slash commands only	claude-flow integrated (60+ agent types, swarm topologies, vector memory)
Copilot support	Partial	Full (.github/copilot-instructions.md + .vscode/settings.json)
Cross-platform guide	No	Yes (CROSS_PLATFORM_INTEGRATION.md)
Shared memory layer	Per-platform	Unified (all 5 platforms read/write the same vector memory)

claude-fulcrum doesn't replace ECC — it extends and integrates it with orchestration, additional language support, and a unified cross-platform layer.

---

🔗 The Cross-Platform Integration

This is the most unique part of this repo. All 5 platforms share a single orchestration layer. Copilot reads patterns learned by Claude Code. Codex workers execute plans made in Cursor. Everything feeds into the same vector memory via claude-flow.

                   claude-fulcrum                         
                                                         
       
   25        108        57         9 Language    
   Agents    Skills     Commands   Rule Sets     
       
                                                     
    
       claude-flow: Shared Vector Memory + MCP         
    
                                                        
    
    Claude Code  Codex CLI  Cursor  Copilot  OC    
    

Each platform reads from and writes to the same shared context:

• Claude Code — deep TDD, planning, security reviews writes patterns to memory
• Codex CLI — fast parallel execution reads plans, writes results
• Cursor — visual IDE integration reads agents/skills, writes code changes
• GitHub Copilot — inline completions reads coding standards and learned patterns
• OpenCode — open-source fallback reads full agent/skill library

Full guide: docs/CROSS_PLATFORM_INTEGRATION.md

---

✨ Features

25 Specialized Agents

Every agent is a domain expert that activates automatically based on context:

Agent	Purpose	Auto-Triggered When
planner	Implementation planning	Complex features, refactoring
architect	System design & scalability	Architectural decisions
tdd-guide	Test-driven development	New features, bug fixes
code-reviewer	Code quality & maintainability	After writing/modifying code
security-reviewer	Vulnerability detection	Before commits, sensitive code
build-error-resolver	Fix build/type errors	Build failures
e2e-runner	End-to-end Playwright testing	Critical user flows
database-reviewer	PostgreSQL/Supabase specialist	Schema design, queries
python-reviewer	Python code review	Python projects
go-reviewer	Go code review	Go projects
kotlin-reviewer	Kotlin/Android/KMP review	Kotlin projects
java-reviewer	Java/Spring Boot review	Java projects
rust-reviewer	Rust code review	Rust projects
cpp-reviewer	C++ code review	C++ projects
doc-updater	Documentation & codemaps	Updating docs
refactor-cleaner	Dead code cleanup	Code maintenance
...and 9 more

108 Workflow Skills

Skills are domain knowledge that activate contextually:

Category	Count	Examples
Languages	30+	TypeScript, Python, Go, Kotlin, Java, Rust, C++, Perl, PHP, Swift
Frameworks	20+	React, Next.js, Django, Spring Boot, Laravel, Ktor, Compose
Testing	15+	TDD workflows, E2E (Playwright), regression, property-based
Architecture	10+	Clean architecture, API design, database patterns, MCP servers
DevOps	10+	Docker, deployment, CI/CD, verification loops
Business	8+	Market research, content engine, investor materials
AI/ML	5+	Continuous learning, autonomous loops, eval harness

57 Slash Commands

/plan          — Implementation planning with task breakdown
/tdd           — Full TDD workflow (RED  GREEN  REFACTOR)
/code-review   — Comprehensive code quality review
/e2e           — Generate and run E2E tests
/build-fix     — Fix build errors automatically
/verify        — Run verification loop (lint + test + typecheck)
/security-scan — Security vulnerability scan
/learn         — Extract reusable patterns from sessions
/skill-create  — Generate new skills from git history
/harness-audit — Audit your harness configuration
/loop-start    — Start autonomous agent loop
...and 46 more

9 Language Rule Sets

Language-specific coding standards enforced automatically:

Rule Set	Focus
Common	Security, error handling, immutability
TypeScript	Strict mode, ESM, React/Next.js patterns
Python	PEP 8, type hints, Pythonic idioms
Go	Standard library first, error handling, testing
Kotlin	Coroutines, null safety, Compose patterns
C++	Modern C++20, memory safety, RAII
Perl	Modern Perl 5.36+, Moose, strict/warnings
PHP	PSR-12, Laravel conventions, Composer
Swift	Swift concurrency, protocol-oriented design

Multi-Agent Orchestration

Coordinated agent swarms via claude-flow:

npx claude-flow@alpha mcp start                              # start MCP server
npx claude-flow@alpha swarm start --topology hierarchical     # run a swarm
npx claude-flow@alpha memory store --key "k" --value "v"      # vector memory

60+ agent types · hierarchical/mesh/ring/star topologies · HNSW vector memory · Raft/BFT consensus · self-learning

---

🖥️ Supported Platforms

Platform	Config Location	Best For
Claude Code	~/.claude/	Deep workflows, TDD, planning
Codex CLI	~/.codex/	Fast parallel execution
Cursor	.cursor/	Visual IDE integration
GitHub Copilot	.github/copilot-instructions.md	Inline completions
OpenCode	.opencode/	Open-source alternative

---

📖 Guides

Guide	Description
Cross-Platform Integration	How all 5 platforms share one orchestration layer — read this first
Extreme Dev Playbook	Daily workflow guide for all platforms working together
Shortform Guide	Setup, foundations, philosophy
Longform Guide	Token optimization, memory persistence, evals, parallelization
Security Guide	Security scanning, secret management, OWASP patterns
Troubleshooting	Common issues and fixes

---

🗂 Architecture

claude-fulcrum/
 agents/              25 specialized agents (planner, reviewer, tdd-guide...)
 skills/              108 workflow skills (TDD, API design, Django, React...)
 commands/            57 slash commands (/plan, /tdd, /code-review, /e2e...)
 hooks/               Hook definitions (pre/post tool use, session events)
 rules/               9 language rule sets
    common/          Universal coding standards
    typescript/      TypeScript/React/Next.js
    python/          Python/Django
    golang/          Go
    kotlin/          Kotlin/Android/KMP
    cpp/             C++
    perl/            Perl
    php/             PHP/Laravel
    swift/           Swift
 mcp-configs/         MCP server configurations
 orchestration/       claude-flow swarm orchestration
 scripts/             Install, CI, hooks, utilities
 docs/                Guides and documentation
 .codex/              Codex CLI platform config
 .cursor/             Cursor platform config
 .opencode/           OpenCode platform config
 .github/             GitHub + Copilot config
 .vscode/             VS Code settings
 install.sh           Unix installer
 install.ps1          Windows installer
 CLAUDE.md            Claude Code project instructions
 AGENTS.md            Agent routing instructions
 package.json         Project manifest

---

🔧 Configuration

MCP Servers

Configured in mcp-configs/:

Server	Purpose
claude-flow	Multi-agent swarm orchestration
memory	Persistent memory across sessions
sequential-thinking	Step-by-step reasoning
context7	Library documentation lookup
playwright	Browser automation & E2E testing
firecrawl	Web scraping & research

Environment Variables

CLAUDE_PACKAGE_MANAGER=pnpm     # Override package manager (npm, pnpm, yarn, bun)
CF_HOOK_PROFILE=standard        # Hook profile (minimal, standard, strict)
CF_DISABLED_HOOKS=cost-tracker  # Comma-separated hook names to disable

---

🧪 Testing

npm test            # Validate agents, commands, rules, skills, hooks
npm run coverage    # Coverage report (80%+ required)
npm run lint        # Lint everything
npm run harness:audit  # Audit harness configuration

---

🤝 Contributing

See CONTRIBUTING.md for details.

1. Fork branch (git checkout -b feat/my-feature) follow formats npm test PR

Formats:

• Agents: Markdown with YAML frontmatter (name, description, tools, model)
• Skills: Markdown with sections (When to Use, How It Works, Examples)
• Commands: Markdown with description frontmatter
• Rules: Markdown in rules/<language>/

File naming: lowercase with hyphens — python-reviewer.md, tdd-workflow.md

---

🏷️ GitHub Topics

When you create this repo on GitHub, add these topics for discoverability:

claude-code · claude-code-agents · ai-dev-workflow · claude-code-setup · codex-cli · cursor-ai · github-copilot · opencode · claude-flow · ai-agents · mcp-server · developer-tools · tdd · code-review · multi-agent · agent-orchestration

---

📊 Stats

Metric	Count
Specialized Agents	25
Workflow Skills	108
Slash Commands	57
Language Rule Sets	9
Supported Platforms	5
MCP Server Configs	6+

---

📜 License

MIT — use it, modify it, share it.

---

The leverage point where all your AI coding tools actually work together.

Back to Top