ofershap/cursor-usage

Overview

This tool provides an MCP server and plugins for Cursor and Claude Code that wrap the full Cursor Enterprise API. It enables AI agents to query and report on team Cursor usage and spending. The server exposes endpoints for retrieving usage data, which can be integrated into agent workflows.

Best for

Best for
Teams using Cursor Enterprise that want to automate spending oversight through AI agents

Use cases

• Monitor team Cursor usage and subscription costs
• Integrate spending data into AI agent queries via MCP
• Automate reporting of Cursor Enterprise usage metrics

How to use

Tools exposed

• get_team_members
• get_spending
• get_daily_usage
• get_billing_groups
• get_usage_events
• set_spend_limit
• get_dau
• get_model_usage
• get_agent_edits
• get_tabs
• get_mcp_usage
• get_file_extensions
• get_client_versions
• get_commands
• get_plans
• get_team_overview
• get_user_deep_dive

Tested with

Cursor, Claude Code

Example client config

{\n  "mcpServers": {\n    "cursor-usage": {\n      "command": "npx",\n      "args": ["-y", "cursor-usage-mcp"],\n      "env": {\n        "CURSOR_API_KEY": "your-api-key-here"\n      }\n    }\n  }\n}

Notes

This tool provides an MCP server and plugins for Cursor and Claude Code that wrap the full Cursor Enterprise API. It enables AI agents to query and report on team Cursor usage and spending. The server exposes endpoints for retrieving usage data, which can be integrated into agent workflows.

3 stars on GitHub. Last updated 2026-02-19. Licensed MIT.

Use cases

• Monitor team Cursor usage and subscription costs
• Integrate spending data into AI agent queries via MCP
• Automate reporting of Cursor Enterprise usage metrics

Pros

• Open source and extensible with standard MCP protocol
• Works seamlessly with Cursor and Claude Code agents
• Provides direct access to the Cursor Enterprise API for spending data

Cons

• Requires a Cursor Enterprise account to function
• Limited to usage and spending queries; no other API features exposed
• Small community and minimal documentation (3 stars on GitHub)

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.