How to Connect Cursor to fanTask via MCP: Step-by-Step Guide

What is the Model Context Protocol (MCP) and how does it work in Cursor?

The Model Context Protocol (MCP) is an open protocol designed to provide AI coding assistants and LLMs with secure, standardized access to external tools, project databases, and custom workflows. Instead of constraining the AI to the boundaries of the open workspace files or manually copy-pasting reference snippets into chat windows, MCP equips the editor with direct, structured APIs that the model can invoke autonomously during code generation and editing.

Within the Cursor editor, native support for the Model Context Protocol enables tools like Cursor Composer and Agent mode to coordinate with external systems while planning, executing multi-file edits, and resolving errors.

Why connect Cursor to fanTask?

Cursor is exceptionally fast and capable at performing broad, multi-file code modifications. However, when working on substantial applications, context windows fill up quickly. Without an external memory layer, Cursor can lose track of foundational architectural decisions, disregard established naming conventions, or unknowingly recreate issues that your team already spent hours debugging.

fanTask provides an MCP-native project ledger and memory system built specifically for AI-assisted software teams. Connecting Cursor to fanTask via MCP delivers key engineering benefits:

• Cursor loads your living architectural guidelines, gotchas, and constraints (via the persistent AI Context field) before initiating major code changes.
• It queries past ticket discussions and resolution logs to check how nuanced edge cases or environment issues were previously solved.
• It synchronizes task status directly on a shared board, documenting progress with structured audit comments referencing specific file paths and line ranges.

Step 1: Get your access token from fanTask

If you just created your account, fanTask automatically shows you a "Connect your AI to fanTask" modal right on your workspace page. Cursor doesn't have a dedicated option in that modal yet, so pick Other — the platform mints a personal access token on the spot and gives you the MCP server URL and authorization header you need for Step 2.

If you closed that modal, or you're on an existing account, you can pick this back up anytime from your workspace menu: Workspace → MCP / AI.

Step 2: Add fanTask to your Cursor MCP Configuration

Cursor configures MCP server integrations using standard JSON formatting. You can access the configuration via Cursor Settings (Features → MCP) or by opening your Cursor MCP configuration file directly (such as ~/.cursor/mcp.json or project-level .cursor/mcp.json).

Add the following configuration block inside the mcpServers section:

{
  "mcpServers": {
    "fantask": {
      "url": "https://api.fantask.ro/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_FANTASK_API_TOKEN"
      }
    }
  }
}

Replace YOUR_FANTASK_API_TOKEN with the personal access token you generated in Step 1. Communication is established over secure TLS directly to the fanTask API endpoint at https://api.fantask.ro/mcp.

Step 3: Restart Cursor and verify available tools

Remember that Model Context Protocol clients load server definitions exclusively upon startup. After saving your MCP configuration file, restart Cursor or trigger the Developer: Reload Window command from the Command Palette to initialize the connection.

Once restarted, Cursor recognizes the fanTask server and makes the following tool suite available in Composer and chat:

• fantask.get_project_context(project_id): Reads the living project documentation, architectural constraints, prohibited dependencies, and mandatory testing rules.
• fantask.search_tasks(query, filters): Performs lexical and semantic searches across historical tickets and comments to avoid repeating past errors.
• fantask.get_task(task_id): Fetches the detailed scope, acceptance criteria, and background context for an individual task.
• fantask.update_task(task_id, status, comment): Updates task progress (e.g. transitioning an item to in_progress or in_review) and records operational notes.
• fantask.create_task(project_id, title, description, priority): Logs newly discovered bugs, tech debt, or follow-up items as dedicated tasks on the board.
• fantask.add_comment(task_id, message): Records structured audit notes outlining changed files, line numbers, and technical justifications.

Recommended Collaborative Workflow with Cursor Composer

Pairing Cursor with fanTask creates an organized, transparent development cycle:

1. Starting a Session: In Cursor Composer, provide your initial prompt: "Check our project AI Context via fanTask and look at active tasks." Cursor invokes get_project_context and search_tasks, grounding its proposals in your team's documented conventions.
2. Feature Development: The assistant transitions the ticket to in_progress using update_task. If it encounters confusing library behavior or an undocumented error, it searches past resolutions using search_tasks.
3. Review & Verification: After completing the implementation and running local test suites, Cursor marks the task as in_review and posts an audit summary via add_comment. You inspect the code diff and mark the task as done in fanTask yourself.

Get Started for Free

The fanTask Free plan gives you 1 full project, up to 3 collaborators, persistent AI Context, and complete MCP tool access. Connect Cursor to fanTask today and bring long-term memory to your AI coding workflow.

Create a free fanTask account and connect Cursor →