4 packages
@marchward/engine
Marchward policy engine — the deterministic decision engine and tamper-evident hash-chain audit primitive for AI agent governance
@marchward/mcp-server
Marchward MCP Server: governance layer for AI coding agents via Model Context Protocol
@marchward/proxy
Marchward proxy — the local runtime governor that intercepts MCP and HTTP agent tool calls, enforces policy, and mediates credentials
@marchward/sdk
Marchward TypeScript SDK — typed client for the runtime authority API for AI agents