// SCAN REPORT

Cursor Cloud Agent

10 tools. 4 can modify or destroy data without limits.

1 destructive tool with no built-in limits. Policy required.

Last updated: 29 Jun 2026

4 can modify or destroy data

6 read-only

10 tools total

Community server · catalogue entry verified 29/06/2026

How to control Cursor Cloud Agent ↓

TOOL MAP

What Cursor Cloud Agent exposes to your agents

Read (6) Write / Execute (3) Destructive / Financial (1)

Running more than one server? Check your whole stack →

MOST DANGEROUS TOOLS

Critical Risk

The most dangerous Cursor Cloud Agent tools

Destructive · High delete_agent Delete a cloud agent. This action is permanent and cannot be undone. Execute · Medium launch_agent Start a new cloud agent to work on your repository Execute · Medium stop_agent Stop a running cloud agent. This pauses the agent\ Write · Medium add_followup Add a follow-up instruction to an existing cloud agent

4 of Cursor Cloud Agent's 10 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

POLICY

How to control Cursor Cloud Agent

PolicyLayer is an MCP gateway — it sits between your AI agents and Cursor Cloud Agent, and nothing reaches the server without passing your rules. These are the rules we recommend:

Deny destructive operations

{
  "delete_agent": {
    "deny_if": [
      {
        "conditions": [],
        "on_deny": "Blocked by default. Requires approval."
      }
    ]
  }
}

Destructive tools should never be available to autonomous agents without human approval.

Rate limit write operations

{
  "add_followup": {
    "limits": [
      {
        "counter": "add_followup_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations

{
  "get_agent": {
    "limits": [
      {
        "counter": "get_agent_per_minute",
        "window": "minute",
        "max": 60,
        "scope": "grant"
      }
    ]
  }
}

Controls API costs and prevents retry loops from exhausting upstream rate limits.

Create a free account and register Cursor Cloud Agent — nothing to install.
Add these rules — paste them, or build them visually. Tune the limits to your setup.
Point your MCP client (Claude, Cursor, anything) at your gateway URL.

ENFORCE POLICY ON CURSOR CLOUD AGENT →

Instant setup, no code required.

FULL CATALOGUE

All 10 Cursor Cloud Agent tools

DESTRUCTIVE 1 tools

Destructive delete_agent Delete a cloud agent. This action is permanent and cannot be undone.

EXECUTE 2 tools

Execute launch_agent Start a new cloud agent to work on your repository Execute stop_agent Stop a running cloud agent. This pauses the agent\

WRITE 1 tools

Write add_followup Add a follow-up instruction to an existing cloud agent

READ 6 tools

Read get_agent Retrieve the current status and results of a cloud agent Read get_agent_conversation Retrieve the conversation history of a cloud agent, including all user prompts and assistant responses Read get_api_key_info Retrieve information about the API key being used for authentication Read list_agents List all cloud agents for the authenticated user Read list_models Retrieve a list of recommended models for cloud agents Read list_repositories Retrieve a list of GitHub repositories accessible to the authenticated user. Note: This endpoint has strict ra

EXPLORE

Related servers

Other MCP servers with similar tools — same risk classification, starter policies for each.

FAQ

Questions about Cursor Cloud Agent

Can an AI agent delete data through the Cursor Cloud Agent MCP server? +

Yes. The Cursor Cloud Agent server exposes 1 destructive tools including delete_agent. These permanently remove resources with no undo. PolicyLayer blocks destructive tools by default so they never reach the upstream server.

How do I prevent bulk modifications through Cursor Cloud Agent? +

The Cursor Cloud Agent server has 1 write tools including add_followup. Set a rate limit in your policy -- for example, 10 calls per hour prevents an agent from making more than 10 modifications per hour. PolicyLayer enforces this at the gateway, before calls reach Cursor Cloud Agent.

How many tools does the Cursor Cloud Agent MCP server expose? +

10 tools across 4 categories: Destructive, Execute, Read, Write. 6 are read-only. 4 can modify, create, or delete data.

How do I enforce a policy on Cursor Cloud Agent? +

Register the Cursor Cloud Agent MCP server in PolicyLayer, apply the suggested rules above (adjust the limits to your use case), and point your AI client at the PolicyLayer proxy URL instead of the server directly. Your agents keep the same tools; PolicyLayer evaluates every call against policy before it executes. Nothing to install, live in minutes.

Enforce policy on every Cursor Cloud Agent tool call.

Deterministic rules across all 10 Cursor Cloud Agent tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

ENFORCE POLICY ON CURSOR CLOUD AGENT →

Instant setup, no code required.

10 Cursor Cloud Agent tools catalogued and risk-classified — across an index of 43,000+ MCP servers.