How to Add Live Market Data to Cursor

To add live market data to Cursor, add the SentiSense MCP server by its URL, https://app.sentisense.ai/mcp, and sign in with a free SentiSense account. If you prefer documented workflows, get a free API key and install the SentiSense skills with one command. Either way, the agent can screen stocks, read sentiment, check analyst coverage and pull insider, institutional and earnings data without leaving the editor.

What is Cursor?

Cursor is an AI code editor: a coding agent that reads your project, runs commands and writes code from a chat panel. For a builder, it is where trading scripts, dashboards and backtests get written. For a trader who codes, it doubles as a research assistant that can turn a question into a script in the same session.

Why does a coding assistant need a market data feed?

On its own, the agent knows markets only as of its training cutoff, months in the past. Asked about a stock, it either guesses or searches the web and repeats whatever headline came back. Connected to a feed, it calls the data itself: the tool returns dated numbers, the agent reads them back, and the code it writes runs against the same source.

How do you set it up?

You need only one of the two paths. Running both is fine.

Path 1: the MCP server (sign in, nothing to paste).

1. Create a free account at app.sentisense.ai. The connector signs in with it.

2. Add the server to ~/.cursor/mcp.json for every project, or to .cursor/mcp.json inside one project:

   {
     "mcpServers": {
       "sentisense": {
         "url": "https://app.sentisense.ai/mcp"
       }
     }
   }
   

3. Open Cursor's MCP settings (Customize in the sidebar) and enable sentisense. When Cursor connects, a browser tab opens on the SentiSense sign-in page. Sign in, approve access, and the browser hands you back to Cursor.

4. Ask a market question in the Agent chat. By default Cursor asks for approval before it runs an MCP tool, and the arrow next to each tool call shows the arguments it sent.

Path 2: the free API key and the skills.

1. Get a free key at sentisense.ai/get-api-key. No card required.

2. From your project directory, run:

   npx skills add SentiSenseApp/skills -a cursor
   

   This installs the skills into .agents/skills/ for the current project. Add -g to install them in ~/.cursor/skills/ for every project. Cursor loads skills from both folders.

3. Put the key in your shell environment as SENTISENSE_API_KEY.

4. Ask a question, or type / in the Agent chat and pick a skill by name.

The connector suits quick questions. The skills carry longer workflows with fixed output templates: a morning dashboard, an earnings review, a position-size check.

What can it actually do?

Five prompts to try once it is connected, with the tool that answers each:

• "What is the market mood today, and which signals are driving it?" get_market_mood returns the SentiSense fear and greed composite with its components.
• "Screen for large caps whose SentiSense Score rose this week." screen_stocks ranks the covered universe by score, mentions, market cap or price change.
• "Who covers [ticker], and which firms moved after the last print?" get_analyst_ratings returns the consensus, price targets and recent rating changes.
• "Which institutions added to [ticker] last quarter?" get_smart_money reads 13F holdings and flows.
• "When does [ticker] report next, and what is consensus EPS?" get_earnings_calendar returns the date, timing and estimate.

Every tool is read-only. The agent can look things up, and nothing more.

The research-then-code pattern

Cursor's edge over a chat app is that research and code share one session. Ask the agent to run a screen through the connector. Then ask it to write the script that reproduces the screen against the REST API with your key, and to rerun it on a schedule or chart the result. The connector answers the question once; the script answers it every morning. The same skills work in Claude Code, Codex and Grok Bot, so the workflow moves with you if your editor does.

Data is provided for research and education, not investment advice.

Get a free API key →