Switch to a different model for the current provider
AI agents use use_model to create or update resources in MemGPT MCP Server — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your MemGPT MCP Server environment.
This is a Write operation because it modifies system state (the active model selection) reversibly. It's not Destructive because the change can be undone by selecting another model. It's not Execute because it doesn't run external code or operations—it only changes a configuration setting.
From the tool's definition The tool switches to a different model for the current provider, which modifies the active configuration state of the MemGPT system.
Documented attack patterns abuse exactly the kind of access use_model gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and MemGPT MCP Server, and nothing reaches the server without passing your rules. This is the rule we recommend for use_model:
{
"version": "1",
"default": "deny",
"tools": {
"use_model": {
"limits": [
{
"counter": "use_model_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}use_model 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.
Switch to a different model for the current provider. It is categorised as a Write tool in the MemGPT MCP Server MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the MemGPT MCP Server MCP server in PolicyLayer and add a rule for use_model: 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 MemGPT MCP Server. Nothing to install.
use_model 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 use_model 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 use_model. 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.
use_model is provided by the MemGPT MCP Server MCP server (vic563/memgpt-mcp-server). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from MemGPT MCP Server, 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.
5 MemGPT MCP Server tools catalogued and risk-classified — across an index of 43,000+ MCP servers.