Explicitly destroy a session, killing its interpreter and wiping its files.
AI agents call destroy_session to permanently remove resources in Code Box — typically in cleanup and lifecycle workflows. It does its job in a single call, and there is no undo.
This tool permanently deletes session data and kills running processes without possibility of recovery. Although scoped to a single session rather than affecting global infrastructure, it represents irreversible data loss (wipe). Severity is high rather than critical because impact is typically limited to one user session, but confidence is very high due to explicit destructive language in both name and description.
From the tool's definition Tool name and description both confirm irreversible deletion: 'destroy' and 'wiping its files' indicate unrecoverable removal of session state and associated data.
Attacks that exploit this kind of access
Explicitly destroy a session, killing its interpreter and wiping its files. It is categorised as a Destructive tool in the Code Box MCP Server, which means it can permanently delete or destroy data. Block by default and require explicit approval.
Register the Code Box MCP server in PolicyLayer and add a rule for destroy_session: 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 Code Box. Nothing to install.
destroy_session is a Destructive tool with critical risk. Critical-risk tools should be blocked by default and only enabled with explicit human approval.
Yes. Add a rate_limit block to the destroy_session 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 destroy_session. 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.
destroy_session is provided by the Code Box MCP server (madhanmohanreddy2301/codeboxmcp). 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 →