The Citadelle MCP server lets AI agents interact with decentralized US stock options markets on EVM, retrieving market liquidity, checking wallet positions, and generating secure trade links.
Fetch options market liquidity: Get all available options markets (e.g., AAPL, TSLA, NVDA) and filter by asset symbol to see active strikes and liquidity.
View wallet portfolio: Retrieve open long and short options positions for any provided EVM wallet address.
Generate secure trade links: Prepare a trade (buy/sell call/put) and output a deep link to the Citadelle web UI for the user to review and sign with their browser wallet.
Provides tools for interacting with Citadelle options markets on the Robinhood Chain, including fetching market liquidity, wallet positions, and generating secure trade links.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@citadelle-mcp What are the current liquid markets on Citadelle?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
An official Model Context Protocol (MCP) server for Citadelle Options - allowing AI agents (like Claude, Cursor, and Windsurf) to securely interact with decentralized US Stock options markets on the EVM / Robinhood Chain.
This MCP server equips AI Assistants with three powerful, context-aware tools:
get_markets_liquidity: Fetches all available US Stock options markets (e.g., AAPL, TSLA, NVDA) from the Citadelle Indexer and filters them by available liquidity. The AI instantly knows which markets are active, what the current strike prices are, and whether there is enough liquidity to execute a trade.
get_wallet_portfolio: Fetches the open Long and Short (Written) US Stock options positions for any provided EVM wallet address.
generate_trade_link: Safely prepares a transaction URL. Instead of having the AI sign a transaction directly with private keys (which is a massive security risk), the AI returns a secure deep-link to the Citadelle Web UI. The user can then review and sign the transaction safely using their browser wallet (MetaMask, Rabby, etc.).
Related MCP server: E*TRADE MCP Server
Once the MCP Server is connected to your AI Assistant, you can use natural language to interact with Citadelle. Try these prompts:
Exploring Markets:
"I want to trade US Stocks on Citadelle. What AAPL options are currently available and have liquidity?" "Show me all available Call options for TSLA."
Checking Portfolio:
"Show me my open options portfolio for wallet 0xYourWalletAddress..." "Do I have any active put options for NVDA in my wallet?"
Executing Trades:
"Buy 100 AAPL Call Options for me." (The AI will find the best market, generate a secure trade link, and provide it to you to click and sign). "I want to write a Put option for TSLA to earn some premium. Generate a transaction link for me."
You can run this server directly via npx in any compatible MCP client without needing to install or clone the repository locally!
Open Settings -> MCP.
Add a new MCP Server.
Name: Citadelle
Type: command
Command: npx
Args: -y citadelle-mcp@latest (This ensures the AI always fetches the newest protocol updates).
Set the Environment Variables within the Cursor UI (see the Environment Variables section below).
Add the following configuration to your claude_desktop_config.json:
{
"mcpServers": {
"citadelle": {
"command": "npx",
"args": ["-y", "citadelle-mcp@latest"],
"env": {
"RPC_URL": "https://api.citadelleoption.tech/api/rpc",
"WEB_UI_BASE_URL": "https://citadelleoption.tech",
"API_BASE_URL": "https://api.citadelleoption.tech/api",
"CONTRACT_ADDRESS": "0xe268a55dD2672Ddb6d7727283B27d9BE601c421a"
}
}
}
}To ensure the AI routes your requests to the correct network and API, you must configure the following environment variables:
Variable | Description | Default / Example |
| Your EVM RPC endpoint (use the Backend Proxy for security) |
|
| The domain of the Citadelle Web UI for deep-links |
|
| The URL of the Citadelle API Indexer |
|
| The EVM address of the Citadelle Options smart contract |
|
If you wish to contribute or run the MCP server locally from source:
Clone the repository and install dependencies:
npm installCreate a .env file from the example:
cp .env.example .envBuild the TypeScript code:
npm run buildTo link it locally to an MCP client like Cursor, point the MCP settings to your local dist/index.js file:
Command: node
Args: /absolute/path/to/citadelle-mcp/dist/index.js
If your AI assistant is returning outdated market data or generating incorrect links, it may be using a cached version of the MCP Server. To force an update:
Claude Desktop: Quit the application completely (Cmd+Q / Ctrl+Q) and reopen it. This forces npx to fetch the @latest version.
Cursor/Windsurf: Restart the MCP Server from the settings panel, or manually run npx clear-npx-cache in your terminal before restarting.
MIT
Generates a URL that redirects the user to the web UI to securely execute an option trade.
| Name | Required | Description | Default |
|---|---|---|---|
| action | Yes | The action to perform: 'buy_call', 'sell_call', 'buy_put', 'sell_put' | |
| marketId | Yes | The EVM address of the Market |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the behavioral burden. It states that the tool only generates a URL and that the trade is executed in the web UI, which rules out direct execution. However, it does not disclose whether authentication is required, whether the link is one-time, or what happens on invalid inputs.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
A single, front-loaded sentence states the action and outcome without fluff. Every word contributes to understanding the tool's purpose.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a simple two-parameter tool with full schema coverage, the description is nearly complete. It specifies the output (a URL), the redirect target, and the trade context. It could add error or authentication behavior, but nothing essential is missing for selecting and invoking it correctly.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the schema already documents marketId and action, including the allowed action values. The description adds only the high-level context of an option trade and does not further clarify parameter semantics; baseline 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a specific verb ('Generates') and a clear resource ('a URL that redirects the user to the web UI to securely execute an option trade'). This is immediately distinguishable from the read-only sibling tools, which fetch market liquidity and wallet portfolio data.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It clearly conveys the use case: when the user needs to initiate an option trade through the web UI, this tool produces the redirect URL. It does not name other tools or give explicit when-not-to-use guidance, but the siblings are sufficiently unrelated that the context alone is clear.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Fetches all Citadelle options markets directly from the EVM blockchain and filters them by available liquidity.
| Name | Required | Description | Default |
|---|---|---|---|
| assetSymbol | No | Optional asset symbol to filter (e.g., SOL, BTC, ETH) |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of behavioral disclosure. It usefully adds that data is fetched 'directly from the EVM blockchain' and that liquidity filtering occurs, but it does not explain what 'available liquidity' means, whether results are paginated, or what the return format is. This is adequate but not rich.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, efficient sentence with no filler or redundancy. It front-loads the core action and resource, then adds the filtering behavior and data source.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a simple read-only listing tool with one optional parameter, the description is mostly sufficient, but it leaves the exact meaning of 'available liquidity' unclear and provides no return-shape information despite there being no output schema. An agent could call it correctly, but might misjudge what results are returned.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema covers 100% of the single parameter assetSymbol, including its optionality and examples. The description itself does not mention this parameter, but the schema already documents it, so the baseline of 3 applies. No additional semantic value is added beyond the schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a specific verb ('Fetches'), a clear resource ('all Citadelle options markets'), and a distinct source ('directly from the EVM blockchain'). It also states the selection criterion ('filters them by available liquidity'), which clearly differentiates this tool from the sibling tools get_wallet_portfolio and generate_trade_link.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies when to use the tool—when market liquidity data is needed—but it does not explicitly mention alternatives or state when not to use it. The sibling tools are clearly different in purpose, but no direct routing guidance is provided.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Fetches open Citadelle options positions for a specific EVM wallet directly from the blockchain.
| Name | Required | Description | Default |
|---|---|---|---|
| walletAddress | Yes | The EVM wallet address (0x string) |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden and does provide some useful behavioral context: it is a read operation ('fetches'), limited to open positions, and sourced directly from the blockchain. However, it does not disclose error behavior, data format, pagination, or whether network conditions or blockchain provider requirements apply.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
A single, front-loaded sentence that conveys the action, resource, scope, and data source with no wasted words. It is appropriately concise for a simple one-parameter tool.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a low-complexity tool with one fully documented parameter, the description gives enough to invoke it correctly and states what the result represents ('open Citadelle options positions'). However, since there is no output schema, return shape and field details are left unspecified, though this is a minor gap for such a simple getter.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema already fully documents the only parameter with 'The EVM wallet address (0x string)', so the baseline is 3. The description's phrase 'specific EVM wallet' reinforces the parameter's meaning but adds no new semantic details beyond the schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description states a specific action ('fetches'), a concrete resource ('open Citadelle options positions'), and an explicit scope ('for a specific EVM wallet'). It is clearly distinguishable from the sibling tools, which are about market liquidity and trade link generation.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies the tool should be used when an agent needs open options positions for a specific EVM wallet, but it does not explicitly state when to use this tool over alternatives or when not to use it. No exclusions or sibling comparisons are provided.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v1.0.3generate_trade_linkget_markets_liquidityget_wallet_portfolioScored across 3 tools
Each tool targets a clearly distinct concern: market liquidity, wallet positions, and trade execution. There is no overlap or plausible confusion between them.
Two tools follow the get_verb_noun pattern and the third uses generate_trade_link, which is a natural and consistent extension of the same verb-object style. All names are in snake_case with clear, readable intent.
Three tools is within the ideal range and each serves a necessary step in the user journey: viewing markets, checking portfolio, and initiating a trade. The scope is compact but well-defined.
The tool set covers the core flow of discovering markets, reviewing positions, and generating a trade link, which is sufficient for most read-and-execute workflows. Minor gaps like fetching individual position details or market history exist but do not create dead ends.
Non-custodial limit, stop-loss and DCA trading on Epsilon (Robinhood Chain) for AI agents
Options analytics for AI assistants: chains, IV rank, VRP, Greeks, GEX, expected moves, screeners.
Unified financial infrastructure connecting AI agents directly to trade live/demo brokerage accounts, Web3 non-custodial wallets, real-time market data across equities, ETFs, crypto, forex, options, DeFi swaps, and prediction markets, institutional research feeds, and algorithmic strategy backtesters.
Non-custodial trading for AI agents: 1,900+ assets — US stocks, treasuries, gold, 250+ perps.
