AI agents call memory_read_graph to retrieve information from VaultAssist without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool is a Read operation that accesses user data at scale ('entire user'). While it has no destructive capability, the severity is elevated to 'high' because it likely retrieves comprehensive personal information across Google Workspace services (emails, documents, calendar events, etc.), representing significant privacy exposure if misused by an untrusted agent.
From the tool's definition Tool name contains 'read' and description states 'Read and retrieve the entire user' - clearly a retrieval operation with no modification capability.
Risk signalsBulk/mass operation — affects multiple targets
Documented attack patterns abuse exactly the kind of access memory_read_graph gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and VaultAssist, and nothing reaches the server without passing your rules. This is the rule we recommend for memory_read_graph:
{
"version": "1",
"default": "deny",
"tools": {
"memory_read_graph": {}
}
}memory_read_graph is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Read and retrieve the entire user. It is categorised as a Read tool in the VaultAssist MCP Server, which means it retrieves data without modifying state.
Register the VaultAssist MCP server in PolicyLayer and add a rule for memory_read_graph: 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 VaultAssist. Nothing to install.
memory_read_graph 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 memory_read_graph 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 memory_read_graph. 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.
memory_read_graph is provided by the VaultAssist MCP server (3xcaffeine/mcp-server). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from VaultAssist, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
79 VaultAssist tools catalogued and risk-classified — across an index of 43,000+ MCP servers.