AI agents call session_info to retrieve information from Code Box without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The 'info' suffix and context of a stateful code execution server strongly suggest this is a query/retrieval operation returning session state or metadata. No description provided limits confidence, but the pattern of sibling read operations (list_artifacts, list_sessions) indicates this follows the same read-only pattern. Blast radius is minimal as it only exposes existing session information.
From the tool's definition Tool named 'session_info' with empty description. Based on naming convention and sibling tools (destroy_session, exec_code, exec_sql, list_artifacts, list_sessions), this appears to retrieve metadata about an execution session.
Attacks that exploit this kind of access
session_info. It is categorised as a Read tool in the Code Box MCP Server, which means it retrieves data without modifying state.
Register the Code Box MCP server in PolicyLayer and add a rule for session_info: 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.
session_info 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 session_info 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 session_info. 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.
session_info 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 →