kernel_pull
AI agents call kernel_pull to retrieve information from Kaggle-MCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Pulling a kernel is fundamentally a read operation that retrieves existing code or notebook content from Kaggle without modifying, executing, or deleting data. While the tool name is somewhat vague due to the empty description, 'pull' in this context most likely means fetching/downloading. Even if it were to execute retrieved code locally, the tool itself does not execute—it only retrieves.
From the tool's definition Tool name 'kernel_pull' suggests retrieving or downloading kernel code/notebooks from Kaggle. The server description indicates it enables 'analyze kernels' and 'download datasets', positioning this as a data retrieval operation.
Attacks that exploit this kind of access
kernel_pull. It is categorised as a Read tool in the Kaggle-MCP MCP Server, which means it retrieves data without modifying state.
Register the Kaggle- MCP server in PolicyLayer and add a rule for kernel_pull: 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 Kaggle-MCP. Nothing to install.
kernel_pull is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the kernel_pull 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 kernel_pull. 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.
kernel_pull is provided by the Kaggle- MCP server (realbytecode/kaggle-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 →