AI agents use save_loaded_notebook to create or update resources in Notebookllm — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Notebookllm environment.
The tool creates or modifies notebook files on disk in a reversible manner (notebooks can be edited again or reverted). While the description is empty, the name and server context provide clear evidence of Write functionality. Severity is high because unintended saves could overwrite important notebook states, and the tool operates on user data files.
From the tool's definition Tool name 'save_loaded_notebook' indicates writing/persisting notebook state to storage. This is confirmed by the server's capability to 'load, edit, and saving notebooks via MCP tools' — saving is explicitly mentioned as a server function.
Documented attack patterns abuse exactly the kind of access save_loaded_notebook gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Notebookllm, and nothing reaches the server without passing your rules. This is the rule we recommend for save_loaded_notebook:
{
"version": "1",
"default": "deny",
"tools": {
"save_loaded_notebook": {
"limits": [
{
"counter": "save_loaded_notebook_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}save_loaded_notebook stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
save_loaded_notebook. It is categorised as a Write tool in the Notebookllm MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Notebookllm MCP server in PolicyLayer and add a rule for save_loaded_notebook: 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 Notebookllm. Nothing to install.
save_loaded_notebook 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 save_loaded_notebook 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 save_loaded_notebook. 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.
save_loaded_notebook is provided by the Notebookllm MCP server (yasirrazaa/notebookllm_mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Notebookllm, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
6 Notebookllm tools catalogued and risk-classified — across an index of 43,000+ MCP servers.