Forgemax

v0.6.0Developer Toolsstable

Collapse multiple servers and tools into two sandboxed MCP tools for efficient, typed JavaScript execution with secure resource access and concurrency.

ai-agentsai-toolsgatewaymcpmcp-gateway
Share:
149
Stars
0
Downloads
0
Weekly
0/5

What is Forgemax?

Forgemax is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to collapse multiple servers and tools into two sandboxed mcp tools for efficient, typed javascript execution with secure resource access and concurrency.

Collapse multiple servers and tools into two sandboxed MCP tools for efficient, typed JavaScript execution with secure resource access and concurrency.

This server falls under the Developer Tools category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • Collapse multiple servers and tools into two sandboxed MCP t

Use Cases

Collapse multiple servers and tools into two sandboxed MCP tools for efficient,
postrv

Maintainer

LicenseNOASSERTION
Languagerust
Versionv0.6.0
UpdatedMay 21, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

NPM

npx -y forgemax

Manual Installation

npx -y forgemax

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Forgemax

Forgemax is a Rust-based MCP gateway that collapses all the tools from multiple downstream MCP servers into exactly two exposed interfaces: `search` (to discover capabilities) and `execute` (to run JavaScript against those capabilities in a sandboxed V8 isolate). This design reduces tool list sprawl in LLM contexts while enforcing strong security boundaries — the sandbox blocks filesystem, network, and environment access through static analysis, V8 isolation, and process-level containment. It supports parallel execution, session-scoped stash storage, and resource access, making it suitable for orchestrating complex multi-server AI agent workflows from a single typed JavaScript execution surface.

Prerequisites

  • Node.js 18+ for npx installation, or Rust toolchain for building from source
  • Forgemax binary installed (via npm, Homebrew, Cargo, or shell script)
  • One or more downstream MCP servers to aggregate (configured in forge.toml)
  • An MCP client: Claude Desktop, Claude Code, or any MCP-compatible host
1

Install Forgemax

Choose your preferred installation method. npm/npx is simplest; Homebrew and Cargo are available for system-wide installs.

# npm (recommended)
npm install -g forgemax

# Homebrew (macOS/Linux)
brew tap postrv/forgemax && brew install forgemax

# Cargo (requires Rust)
cargo install forgemax

# Shell script (macOS/Linux)
curl -fsSL https://raw.githubusercontent.com/postrv/forgemax/main/install.sh | bash
2

Initialize the configuration

Run `forgemax init` in your project directory to generate a `forge.toml` configuration file with example server definitions.

forgemax init
3

Configure downstream MCP servers in forge.toml

Edit `forge.toml` to list each MCP server you want Forgemax to aggregate. Each server entry needs a command, optional args, and a transport type.

[servers.filesystem]
command = "npx"
args = ["-y", "@modelcontextprotocol/server-filesystem", "/tmp"]
transport = "stdio"

[servers.github]
command = "npx"
args = ["-y", "@modelcontextprotocol/server-github"]
transport = "stdio"

[sandbox]
timeout_secs = 5
max_heap_mb = 64
execution_mode = "child_process"
4

Add Forgemax to your MCP client config

Register Forgemax as the single MCP server your client connects to. It will aggregate all servers defined in forge.toml and expose them through the two gateway tools.

{
  "mcpServers": {
    "forgemax": {
      "command": "npx",
      "args": ["-y", "forgemax"]
    }
  }
}
5

Use the search tool to discover capabilities

In your MCP client, call the `search` tool with a natural language query to find available tools across all configured downstream servers. No credentials are exposed to the sandbox.

6

Execute JavaScript in the sandbox

Use the `execute` tool to run typed JavaScript that orchestrates tool calls across servers. The sandbox API includes forge.callTool, forge.readResource, forge.stash, and forge.parallel.

// Example: parallel file reads with stashing
const schema = await forge.readResource("db", "postgres://mydb/tables");
await forge.stash.put("db_schema", schema, { ttl: 3600 });

const { results } = await forge.parallel([
  () => forge.callTool("filesystem", "read_file", { path: "a.ts" }),
  () => forge.callTool("filesystem", "read_file", { path: "b.ts" })
]);

Forgemax Examples

Client configuration (Claude Desktop)

Minimal Claude Desktop config that launches Forgemax via npx. All downstream server configuration lives in forge.toml in your working directory.

{
  "mcpServers": {
    "forgemax": {
      "command": "npx",
      "args": ["-y", "forgemax"]
    }
  }
}

Prompts to try

Example Claude prompts that leverage the Forgemax gateway's search and execute tools.

- "Search for tools related to reading files across all connected servers"
- "Execute JavaScript to read two files in parallel and compare their contents"
- "Use forge.stash to cache the database schema and then query it in a follow-up execution"
- "List all available tools from the downstream servers via the search capability"
- "Run a sandboxed script that calls the GitHub server to list open PRs and the filesystem server to save the results"

Troubleshooting Forgemax

Sandbox execution is rejected with a static analysis error

The oxc static analyzer blocks `import`, `require`, and `eval` inside sandbox scripts. Rewrite your script to use only the `forge.*` API surface. Network and filesystem calls must go through forge.callTool or forge.readResource, never directly.

Downstream server tools are not appearing in search results

Verify the server command and args in forge.toml are correct and the executable is on your PATH. Run each downstream server command manually in a terminal to check it starts without errors before adding it to forge.toml.

Execution times out before completing

Increase the `timeout_secs` value in the `[sandbox]` section of forge.toml. The default is 5 seconds. For long-running operations, also consider increasing `max_heap_mb` if the script processes large data sets.

Frequently Asked Questions about Forgemax

What is Forgemax?

Forgemax is a Model Context Protocol (MCP) server that collapse multiple servers and tools into two sandboxed mcp tools for efficient, typed javascript execution with secure resource access and concurrency. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Forgemax?

Install via npm with the command: npx -y forgemax. Then add the server configuration to your AI client's JSON config file (e.g., claude_desktop_config.json or .cursor/mcp.json).

Which AI clients work with Forgemax?

Forgemax works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Forgemax free to use?

Yes, Forgemax is open source and available under the NOASSERTION license. You can use it freely in both personal and commercial projects.

Browse More Developer Tools MCP Servers

Explore all developer tools servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "forgemax": { "command": "npx", "args": ["-y", "forgemax"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Forgemax?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides