Manually save a reasoning session with an optional custom name.
AI agents use deep_save_session to create or update resources in Deep Reasoning — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Deep Reasoning environment.
This tool writes/creates a new record (session state) and can be reversed by deleting the session. It has no destructive capability, does not execute external code or commands, and does not move money. The blast radius of misuse is minimal—an agent could save unwanted sessions, but this is easily remedied. Severity is low because the operation is non-destructive and the impact is confined to session storage.
From the tool's definition Tool performs 'save' operation with optional custom naming, which creates or persists data (a reasoning session) in a reversible manner. Description explicitly states 'save' and allows optional naming, consistent with Write category operations.
Attacks that exploit this kind of access
Manually save a reasoning session with an optional custom name. It is categorised as a Write tool in the Deep Reasoning MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Deep Reasoning MCP server in PolicyLayer and add a rule for deep_save_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 Deep Reasoning. Nothing to install.
deep_save_session 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 deep_save_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 deep_save_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.
deep_save_session is provided by the Deep Reasoning MCP server (mauriziomocci/deep-reasoning-mcp). 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 →