AI agents call session_get to retrieve information from Atlas without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves session metadata (state, conversation history, summaries) without creating, modifying, or deleting data. It has minimal blast radius as it only returns information about an existing session. The tool operates on session state that is already local to the MCP server and poses no risk of unintended side effects, data destruction, or external operations.
From the tool's definition Tool name is 'session_get' and description states 'Get session state including recent turns and summary' — uses retrieval verb 'Get' with no modification or deletion indicated.
Attacks that exploit this kind of access
Get session state including recent turns and summary. It is categorised as a Read tool in the Atlas MCP Server, which means it retrieves data without modifying state.
Register the Atlas MCP server in PolicyLayer and add a rule for session_get: 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 Atlas. Nothing to install.
session_get 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_get 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_get. 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_get is provided by the Atlas MCP server (rsanandres/atlas_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 →