🚀 @h0wzy/mcp — The Ultimate Multi-Agent MCP Hub & Go CLI

Centralize, enhance, and distribute high-performance MCP (Model Context Protocol) servers connecting the world's leading AI developer CLIs:

• Claude Code (Anthropic)
• OpenAI Codex CLI (GPT-6 family, default GPT-6-Astra)
• Google Antigravity (Gemini, default Gemini 3.8 Flash)

Cross-model code reviews, independent second opinions, and autonomous multi-agent validation — configured effortlessly via an interactive Golang TUI CLI.

⚡ Instant Run (Zero Setup Required)

Run directly with npx — no pre-installation of Go or local binaries required! The runner automatically retrieves the native binary for your OS and architecture:

npx @h0wzy/mcp

Or install globally via npm (provides hmcp, hwzmcp, and h0wzy-mcp commands):

npm install -g @h0wzy/mcp
hmcp

🎮 CLI Features & Commands

# Launch interactive TUI configuration
hmcp

# Check version and verify if updates are available
hmcp version

# Upgrade hmcp to latest release (or 'hmcp update')
hmcp upgrade

# Setup PATH and shims (~/.local/bin) so 'hmcp' works everywhere
hmcp setup-path

# Diagnose local environment, CLI installations, and paths
hmcp doctor

# Diagnose and output as JSON ({ tools, bridges, cycles, guard })
hmcp doctor --json

# Install all supported integrations automatically
hmcp install --all

# Install a specific bridge globally or per-project
hmcp install claude-antigravity --scope user
hmcp install claude-codex --scope project
hmcp install codex-claude            # let Codex call Claude Code
hmcp install antigravity-claude      # let Antigravity call Claude Code

# List available bridges
hmcp list

# Remove an integration
hmcp remove claude-antigravity

🔌 Available MCP Tools (Symmetrical Parity)

| Provider | Tool Name | Description | | :--- | :--- | :--- | | Antigravity | ask_antigravity | Independent second opinion or general inquiry from Gemini (3.8 Flash by default) | | Antigravity | review_antigravity | Comprehensive code & security review inspecting correctness, edge cases, and diffs | | Antigravity | brainstorm_antigravity | Architectural exploration, trade-offs, and design patterns with Gemini | | Antigravity | plan_antigravity | Structured implementation roadmaps and dependency-ordered execution steps | | Antigravity | delegate_antigravity | Edits files. Hands an implementation task to Antigravity inside cwd (prompt, cwd, optional paths, model, effort, timeout_minutes 1-60, default 30). Review the diff afterwards | | Codex | ask_codex | Cross-verification with OpenAI Codex (GPT-6-Astra by default). Read-only sandbox | | Codex | review_codex | Structured repository code review from OpenAI Codex | | Codex | brainstorm_codex | Architectural exploration, trade-offs, and system design ideation with Codex | | Codex | plan_codex | Step-by-step implementation planning and checklist generation | | Codex | delegate_codex | Edits files. Hands an implementation task to Codex inside cwd (workspace-write sandbox; prompt, cwd, optional paths, model, effort, timeout_minutes 1-60, default 30). Review the diff afterwards |

Delegation tools write to disk. delegate_codex and delegate_antigravity run the agent inside the given cwd and may create or modify files. They never commit for you: always review the resulting diff (git diff) before keeping the changes.

Defaults: Antigravity tools use Gemini 3.8 Flash (High) and Codex tools use gpt-6-astra at medium. Change them per call (model, effort), per session (configure_antigravity / configure_codex), or at startup (AGY_MODEL / AGY_EFFORT, CODEX_MODEL / CODEX_EFFORT).

🗺️ Roadmap / Future

Idea, not implemented. A "super orchestrator" where Claude, Codex and Antigravity share a single task list (for example GitHub Projects) and exchange messages with each other, so they can pick up, hand off and review tasks without a human relaying every step. Today they are only driven one-way through the MCP tools above.

🔒 Security & Privacy

• Automatic Token Redaction: Built-in regex sanitizers proactively scrub OpenAI keys, Google Gemini keys, GitHub/NPM tokens, and Bearer authorization headers before errors or diagnostic messages reach host agents.
• OIDC Trusted Publishing: Distributed directly via GitHub Actions using OpenID Connect (OIDC) and SLSA Provenance attestations with zero static tokens.
• Owner-Only File Permissions: Configuration files (.claude.json, .codex/config.toml, mcp_config.json) are secured with mode 0600 (0700 for directories) on Unix systems.

📄 Repository & Full Documentation

For interactive system architecture diagrams, source code, and release notes:
👉 GitHub Repository: H0wZy/mcp