MCP & A2A Protocols

VORTΞXHQ ships with both a Model Context Protocol (MCP) client/server and an Agent-to-Agent (A2A) protocol so external tools and agents can plug straight in.

Model Context Protocol (MCP)

MCP is the open standard for exposing tools and resources to LLMs. VORTΞXHQ provides:

MCP Client — connect to any MCP server (filesystem, GitHub, Postgres, Slack, custom). Tools become available to every agent.
MCP Server — VORTΞXHQ itself exposes tools (SQL query, SSH exec, API send, FTP read) so external IDEs and agents can drive the app remotely.

Manage servers in Settings → AI → MCP Servers: add, connect, refresh tools, view tool inventories.

Agent-to-Agent (A2A)

A2A lets agents discover each other, exchange task cards, and delegate work — locally or across the network. Use cases:

A specialist on your laptop hands off a deploy task to the DevOps agent on a build server.
A teammate's agent picks up a code-review task you sent through cloud sync.
External A2A-compatible agents (Gemini agents, custom Python agents) join VORTΞXHQ workflows.

Agent Cards

Every VORTΞXHQ agent publishes a card describing its skills, tool inventory, accepted inputs, and SLA hints. Other agents discover capabilities at runtime — no hardcoding.