What is MCP (Model Context Protocol)?
MCP (Model Context Protocol) is an open standard, created by Anthropic, that lets AI applications connect to external tools and data in one permission-based format, and Pluang Agentic Trading is built on it. Instead of a custom integration for every pairing of AI assistant and app, MCP gives developers one shared format for exposing an app's data and actions to an AI agent. The AI can then read information and, where you allow it, take actions, but only within the permissions you grant, which the app enforces.
The parts of an MCP connection, and what they are on Pluang
| Part | What it is | On Pluang |
|---|---|---|
| Server | An app that exposes specific data and actions as MCP "tools" | Pluang's MCP server, which offers tools such as checking holdings, looking up prices and placing orders |
| Client | An AI application that can discover and call those tools | An AI app such as Claude, ChatGPT or Codex |
| Permissions | What the user allows the AI agent to see and do | Read access and trade access are granted separately, on Pluang's own login page |
| Open standard | Any MCP-compatible client can in principle connect to any MCP-compatible app | Pluang follows the shared specification rather than a private protocol |
MCP itself is not specific to trading or to Pluang; it is a general connector standard used across many kinds of software. On Pluang, it is the layer through which a connected AI agent checks your portfolio, looks up prices for US Stocks, Indonesian Stocks and Crypto, reviews transaction history, and places orders on your instructions.
Learn more: Pluang Agentic Trading and what Agentic Trading on Pluang is.
Related questions:
Q: Did Pluang create MCP?
No. MCP was created by Anthropic as an open standard. Pluang adopted it to build Agentic Trading, implementing its own MCP server that follows the shared specification.
Q: Is MCP the same thing as Pluang Agentic Trading?
No. MCP is the general connection standard; Pluang Agentic Trading is Pluang's product built on it, which defines the specific actions a connected AI agent can take on your account, such as checking balances and placing trades.
Q: Do I need to understand MCP to use Pluang Agentic Trading?
No. MCP works behind the scenes, and using Agentic Trading feels like a conversation with your AI agent. You only add Pluang's MCP server URL to your AI app and approve the connection.
Q: Can any AI assistant use MCP to connect to Pluang?
Only AI apps that support custom MCP connections, such as Claude, ChatGPT and Codex. Even then, the connection works only after you explicitly link your Pluang account and approve it on Pluang's login page.
Q: How does MCP limit what an AI agent can access?
MCP is built around explicit, scoped permissions: you grant access to specific data and actions, and Pluang enforces those limits on its side. Orders placed through a connected agent go through the same licensed companies and asset protections as manual trades.
Last updated: 6 November 2026