anvil

<p align="center"> <picture> <source media="(prefers-color-scheme: dark)" srcset="website/public/banner.svg"> </picture> </p>

<p align="center"> <a href="https://www.npmjs.com/package/@anvil-tools/cli"></a> <a href="https://github.com/64envy64/anvil/actions"></a> <a href="https://github.com/64envy64/anvil/blob/main/LICENSE"></a> <a href="https://anvil-sooty.vercel.app"></a> </p>

<p align="center"> <strong>The universal tool compiler for AI agents.</strong><br /> One YAML definition. Ten compilation targets. Zero drift. </p>

<br />

<p align="center"> </p>

---

Every agent runtime has its own tool format. You end up maintaining:

• An MCP schema for Claude Desktop
• An OpenAPI spec for your REST API
• TypeScript types for your SDK
• Hand-written docs that drift out of sync
• No eval coverage, no permission model, no agent-specific descriptions

Anvil replaces all of them with a single source of truth.

npm install -g @anvil-tools/cli
anvil init my-tools && cd my-tools
anvil compile --target mcp            # generate MCP server
anvil serve --stub tools.anvil.yaml   # or run directly as MCP

<p align="center"> </p>

Use in Claude Desktop — zero codegen needed:

{
  "mcpServers": {
    "my-tools": {
      "command": "npx",
      "args": ["@anvil-tools/cli"