subgraph

THE RISK

Low Risk

Why subgraph needs a policy

The tool extracts (reads/queries) a portion of the graph based on specified nodes. It performs no modifications, deletions, or executions—purely a data retrieval operation. Fuzzy label matching is a query feature, not a write operation. Severity is low as misuse can only expose graph data, not alter it.

From the tool's definition "Extract subgraph with specified nodes and their interconnections. Fuzzy matches labels."

Documented attack patterns abuse exactly the kind of access subgraph gives an agent:

POLICY

How to control subgraph

PolicyLayer is an MCP gateway — it sits between your AI agents and Mcp Graph Engine, and nothing reaches the server without passing your rules. This is the rule we recommend for subgraph:

policy.json

{
  "version": "1",
  "default": "deny",
  "tools": {
    "subgraph": {}
  }
}

subgraph is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.

Create a free account and register Mcp Graph Engine — nothing to install.
Add this policy — paste it, or build it visually.
Point your MCP client (Claude, Cursor, anything) at your gateway URL.

CAP THIS TOOL →

Free to start. No card required.

EXPLORE

Related tools and policies

More Mcp Graph Engine tools

Destructive delete_graph Delete a graph session Destructive forget Remove a node and its connected edges Destructive forget_relationship Remove an edge from the graph Destructive transitive_reduction Remove redundant transitive edges (if A->B->C and A->C exist, remove A->C) Execute stop_visualization Stop the visualization server. Execute cypher_query cypher_query Execute visualize_graph Open browser visualization at localhost URL. Updates in real-time. Write add_facts Add facts (relationships) to the graph. Nodes are auto-created if they don

All 24 Mcp Graph Engine tools →

Read tools on other servers

M-Team MCP Server get_torrent_detail YouTube MCP Server search_channel Android Forensics ADB MCP Server parse_browser_history 0xarchive web3_challenge

Go deeper

The MCP Attack Database → Documented attack patterns against MCP deployments — and the policies that stop them.
Data exfiltration → How read access becomes an exfiltration channel when chained with untrusted content.
MCP token cost → What connecting this server costs in context-window tokens on every request.
Rate Limiting MCP Tool Calls: A Practical Guide →
MCP Security: Why Prompt Guardrails Aren't Enough →

FAQ

Questions about subgraph

What does the subgraph tool do? +

Extract subgraph with specified nodes and their interconnections. Fuzzy matches labels. It is categorised as a Read tool in the Mcp Graph Engine MCP Server, which means it retrieves data without modifying state.

How do I enforce a policy on subgraph? +

Register the Mcp Graph Engine MCP server in PolicyLayer and add a rule for subgraph: 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 Mcp Graph Engine. Nothing to install.

What risk level is subgraph? +

subgraph is a Read tool with low risk. Read-only tools are generally safe to allow by default.

Can I rate-limit subgraph? +

Yes. Add a rate_limit block to the subgraph 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.

How do I block subgraph completely? +

Set action: deny in the PolicyLayer policy for subgraph. 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.

What MCP server provides subgraph? +

subgraph is provided by the Mcp Graph Engine MCP server (utilitydelta/mcp-graph-engine). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.

Enforce policy on every Mcp Graph Engine tool call.

Start from Mcp Graph Engine, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.

CHECK YOUR STACK →