AI agents use pubsub.create-topic to create or update resources in Mcp — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Mcp environment.
The tool creates a new resource (a pubsub topic) owned by the user's wallet. This is a Write operation because it creates data reversibly—topics can typically be deleted or modified. Severity is medium because scope is limited to the caller's wallet (no cross-user blast radius), but misconfiguration could lead to unintended topic proliferation or disruption of messaging patterns.
From the tool's definition Tool description states 'create a topic', which is a reversible creation operation that modifies state within the caller's wallet scope.
Attacks that exploit this kind of access
PUBSUB: create a topic you own, scoped to YOUR wallet. It is categorised as a Write tool in the Mcp MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the MCP server in PolicyLayer and add a rule for pubsub.create-topic: 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 Mcp. Nothing to install.
pubsub.create-topic 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 pubsub.create-topic 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 pubsub.create-topic. 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.
pubsub.create-topic is provided by the MCP server (@2sio/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 →