Step 1 10 min

MOSS + Model Context Protocol

Register MCP servers and govern what tools your AI agents can call.

MCP Server Governance

Control which MCP tools your agents can access and under what policies.

The Model Context Protocol (MCP) lets AI agents call external tools. MOSS governs MCP servers: you register servers, create tool-level policies (allow/block specific tool patterns), and every tool call is signed and audited.

Register an MCP Server

Register MCP Servertypescript
await fetch('https://api.mosscomputing.com/v1/mcp/servers', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer ' + access_token,
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    name: 'GitHub Tools',
    url: 'https://mcp.github.example.com',
    transport: 'sse',
    tool_patterns: ['github.*'],
  }),
});

Create MCP Tool Policies

MCP Policytypescript
await fetch('https://api.mosscomputing.com/v1/mcp/policies', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer ' + access_token,
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    mcp_server_id: 'server_abc123',
    tool_pattern: 'github.delete.*',
    action: 'block',
    priority: 10,
  }),
});

MCP Integration complete!

  • Register MCP servers with MOSS
  • Create tool-level policies (block delete, allow read)
  • Every MCP tool call is signed and audited