Delete documents by their IDs
AI agents call delete_document to permanently remove resources in Lightrag — typically in cleanup and lifecycle workflows. It does its job in a single call, and there is no undo.
This tool permanently removes documents from the system with no indication of undo/recovery capability. Deletion is irreversible and constitutes data destruction. While the blast radius is limited by requiring specific document IDs (not a blanket deletion), the permanent nature and context within a knowledge/graph system (LightRag) where documents are core assets makes this high severity.
From the tool's definition Tool name is 'delete_document' with description 'Delete documents by their IDs' — explicitly performs irreversible deletion of data.
Attacks that exploit this kind of access
Delete documents by their IDs. It is categorised as a Destructive tool in the Lightrag MCP Server, which means it can permanently delete or destroy data. Block by default and require explicit approval.
Register the Lightrag MCP server in PolicyLayer and add a rule for delete_document: 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 Lightrag. Nothing to install.
delete_document is a Destructive tool with critical risk. Critical-risk tools should be blocked by default and only enabled with explicit human approval.
Yes. Add a rate_limit block to the delete_document 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 delete_document. 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.
delete_document is provided by the Lightrag MCP server (lightrag-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 →