create_task

THE RISK

High Risk

Why create_task needs a policy

This tool triggers an autonomous agent to execute work on a GitHub repository, which can include code changes and pull request generation. The effects depend on the task prompt arguments and are external operations on real repositories. While not directly destructive or financial, it executes potentially wide-ranging operations autonomously, warranting high severity.

From the tool's definition "Launch a new cloud task to work on a repository" and "Cursor Cloud Agents that autonomously work on GitHub repositories, including creating tasks, monitoring progress, and automatically generating pull requests"

Documented attack patterns abuse exactly the kind of access create_task gives an agent:

POLICY

How to control create_task

PolicyLayer is an MCP gateway — it sits between your AI agents and Cloud Agent MCP Server, and nothing reaches the server without passing your rules. This is the rule we recommend for create_task:

policy.json

{
  "version": "1",
  "default": "deny",
  "tools": {
    "create_task": {
      "limits": [
        {
          "counter": "create_task_rate",
          "window": "minute",
          "max": 10,
          "scope": "grant"
        }
      ]
    }
  }
}

create_task stays usable, but rate-capped — a runaway agent can't fire it dozens of times a minute. Everything else on the server is denied unless you say otherwise.

Create a free account and register Cloud Agent MCP Server — nothing to install.
Add this policy — paste it, or build it visually.
Point your MCP client (Claude, Cursor, anything) at your gateway URL.

RATE-LIMIT THIS TOOL →

Free to start. No card required.

EXPLORE

Related tools and policies

More Cloud Agent MCP Server tools

Destructive delete_task Permanently delete a cloud task. This action cannot be undone and all conversation history Execute add_followup Send additional instructions to a RUNNING task. Use this to guide the task, request change Read get_conversation Get the complete conversation history of a task including the original prompt, all follow- Read get_me Get API key information including name, creation date, and user email. Use this to verify Read get_models List all available LLM models for cloud tasks. If you omit the model parameter in \ Read get_repos Get available repositories. First checks if you are in a git directory and returns that re Read get_task Get the current status and full details of a specific cloud task. Returns comprehensive in Read list_tasks List all cloud tasks for the authenticated user. Returns comprehensive information includi

All 9 Cloud Agent MCP Server tools →

Execute tools on other servers

Android Forensics ADB MCP Server adb_shell_command EMBA-MCP run_emba_scan Frida Game Hacking MCP remove_breakpoint Procmon request_elevation

Go deeper

High-risk tools in Cloud Agent MCP Server → Every high-risk tool this server exposes, in one view.
All high-risk MCP tools → Every high-risk tool across the catalogue, and the policy pattern that contains them.
The MCP Attack Database → Documented attack patterns against MCP deployments — and the policies that stop them.
Agent sandbox → Bounding what an agent can execute and where the boundary should sit.
Least privilege for MCP → Granting agents the minimum tool surface the task needs.
Rate Limiting MCP Tool Calls: A Practical Guide →
MCP Security: Why Prompt Guardrails Aren't Enough →

FAQ

Questions about create_task

What does the create_task tool do? +

Launch a new cloud task to work on a repository. Requires a repository URL and task prompt. Returns a task ID that you can use to monitor progress with \. It is categorised as a Execute tool in the Cloud Agent MCP Server MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

How do I enforce a policy on create_task? +

Register the Cloud Agent MCP Server MCP server in PolicyLayer and add a rule for create_task: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Cloud Agent MCP Server. Nothing to install.

What risk level is create_task? +

create_task is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.

Can I rate-limit create_task? +

Yes. Add a rate_limit block to the create_task rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.

How do I block create_task completely? +

Set action: deny in the PolicyLayer policy for create_task. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.

What MCP server provides create_task? +

create_task is provided by the Cloud Agent MCP Server MCP server (jxnl/cursor-cloud-agent-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.

Enforce policy on every Cloud Agent MCP Server tool call.

Start from Cloud Agent MCP Server, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.

CHECK YOUR STACK →