AI agents invoke create-container to trigger actions in Docker. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call — builds kicked off, notifications sent, workflows started.
Creating a Docker container executes a new isolated environment on the host, which can run arbitrary workloads, expose ports, mount volumes, and consume resources. This goes beyond a simple write operation because it triggers actual process execution. Misuse could lead to running malicious images, resource exhaustion, or network exposure. Severity is high given the blast radius of arbitrary container execution.
From the tool's definition "Create a new standalone Docker container" — instantiates and runs a new container process on the host system
Attacks that exploit this kind of access
Create a new standalone Docker container. It is categorised as a Execute tool in the Docker MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Docker MCP server in PolicyLayer and add a rule for create-container: 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 Docker. Nothing to install.
create-container is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the create-container 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-container. 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-container is provided by the Docker MCP server (quantgeekdev/docker-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 →