Go back in browser history.
AI agents invoke back to trigger actions in ComputerMate. 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.
Navigating back in browser history is an external operation that changes the browser's state. It's not purely reading data, and while reversible in theory, it can trigger page loads, form resubmissions, or other side effects. In the context of an AI controlling a computer, this is an Execute-level action with medium severity since it could navigate away from important pages or disrupt workflows.
From the tool's definition Go back in browser history — triggers a browser navigation action, an external operation with side effects depending on current browser state.
Attacks that exploit this kind of access
Go back in browser history. It is categorised as a Execute tool in the ComputerMate MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the ComputerMate MCP server in PolicyLayer and add a rule for back: 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 ComputerMate. Nothing to install.
back 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 back 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 back. 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.
back is provided by the ComputerMate MCP server (one710/computermate). 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 →