AI agents call read_nodes to retrieve information from Pen without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves document structure and node properties without side effects. It queries the document tree with optional filtering by node IDs and depth, returning data only. No creation, modification, deletion, or external execution occurs. Blast radius is minimal — worst case an agent reads data it shouldn't, but cannot alter or destroy anything.
From the tool's definition Tool name is 'read_nodes' and description states 'Read nodes from the document with depth control' — explicit read operation with no modification capability.
Attacks that exploit this kind of access
Read nodes from the document with depth control. Omit nodeIds to get all top-level children of the page. Use depth=0 for node properties only (children truncated to. It is categorised as a Read tool in the Pen MCP Server, which means it retrieves data without modifying state.
Register the Pen MCP server in PolicyLayer and add a rule for read_nodes: 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 Pen. Nothing to install.
read_nodes 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 read_nodes 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 read_nodes. 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.
read_nodes is provided by the Pen MCP server (@zseven-w/pen-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 →