MCP Servers & A2A Protocol

Plug Anthropic Model Context Protocol servers and Agent-to-Agent endpoints into your VORTΞXHQ agents — extend any agent with custom tools without writing code.

Model Context Protocol (MCP)

MCP is an open standard for exposing tools to LLMs. VORTΞXHQ acts as both an MCP client (consuming external servers) and an MCP server (exposing its modules to other clients).

Add an MCP server

Open Settings → AI → MCP Servers.
Click Add server.
Choose transport: stdio (local executable) or SSE (HTTP endpoint).
Enter the command + args, or the URL + headers.
Click Test — VORTΞXHQ pings the server, lists discovered tools, and previews their schemas.
Save. The tools become available to every agent.

Per-agent tool selection

Each agent's settings panel shows the master tool list grouped by source (built-in · MCP server name · A2A peer). Untick the ones you don't want this agent to call.

Agent-to-Agent (A2A)

The A2A protocol lets VORTΞXHQ delegate sub-tasks to a remote agent endpoint. Use it to:

Send specialised work to a private internal agent (e.g., a fine-tuned company-specific model).
Federate workloads across teammates.
Compose pipelines where each step lives in a different runtime.

A2A configuration

{
  "name": "internal-research",
  "endpoint": "https://agents.acme.dev/research",
  "auth": { "type": "bearer", "token": "{{vault.acme_a2a}}" }
}

Exposing VORTΞXHQ as an MCP server

Enable Settings → AI → Expose as MCP server to let external clients (Claude Desktop, Cursor, custom scripts) call your local SSH/SQL/API/FTP tools through the MCP standard. Authentication is mandatory; tokens are issued per client.