AI agents use create_collection to create or update resources in Metabase — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Metabase environment.
Creating a collection in Metabase is a reversible write operation that adds a new organizational container for storing queries, dashboards, and models. While it modifies the Metabase system state, it does not execute code, run queries, or delete data. An AI agent misusing this could create numerous collections cluttering the system, but the impact is reversible via deletion.
From the tool's definition Tool name 'create_collection' and description 'Create a new collection in Metabase' indicate data creation.
Attacks that exploit this kind of access
Create a new collection in Metabase. It is categorised as a Write tool in the Metabase MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Metabase MCP server in PolicyLayer and add a rule for create_collection: 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 Metabase. Nothing to install.
create_collection 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_collection 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_collection. 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_collection is provided by the Metabase MCP server (voducdan/matebase-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 →