AI agents call hashgnn to retrieve information from Neo4j Gds without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves/computes derived properties (node embeddings) from the graph without side effects. It does not modify, delete, or execute external operations. It fits the Read category: a query-like operation that extracts or computes graph-based information for analysis. The low severity reflects that embedding computations have no blast radius beyond potentially incorrect analytical results.
From the tool's definition HashGNN is described as a 'node embedding algorithm' that 'does not include a model or require training.' It computes embeddings from graph structure, consistent with graph analysis and pattern discovery.
Attacks that exploit this kind of access
HashGNN is a node embedding algorithm which resembles Graph Neural Networks but does not include a model or require training. It is categorised as a Read tool in the Neo4j Gds MCP Server, which means it retrieves data without modifying state.
Register the Neo4j Gds MCP server in PolicyLayer and add a rule for hashgnn: 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 Neo4j Gds. Nothing to install.
hashgnn 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 hashgnn 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 hashgnn. 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.
hashgnn is provided by the Neo4j Gds MCP server (neo4j-contrib/gds-agent). 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 →