AI agents use create_subtask to create or update resources in Smokeball — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Smokeball environment.
This tool creates a new subtask record within a law firm practice management system, which is a Write operation (reversible data creation). The severity is medium because misuse could create unwanted task hierarchy clutter affecting workflow organization, but the impact is limited to task metadata rather than financial data, case-critical information, or irreversible deletions.
From the tool's definition Tool name is 'create_subtask' and description states 'Create a subtask under a task.' The verb 'create' indicates data creation, and subtasks are reversible organizational entities that can be modified or deleted.
Attacks that exploit this kind of access
Create a subtask under a task. It is categorised as a Write tool in the Smokeball MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Smokeball MCP server in PolicyLayer and add a rule for create_subtask: 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 Smokeball. Nothing to install.
create_subtask is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the create_subtask 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.
Set action: deny in the PolicyLayer policy for create_subtask. 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.
create_subtask is provided by the Smokeball MCP server (rosenadvertising/smokeball-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →