AI agents use merge_entities to create or update resources in Lightrag — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Lightrag environment.
The merge_entities tool creates a modified state by consolidating entities and their relationships. While relationships are preserved (suggesting reversibility in principle), the operation fundamentally rewrites entity records and their interconnections in the knowledge graph. This is more severe than a simple Read operation but does not permanently destroy data or prevent undo (it's not Destructive).
From the tool's definition Tool performs merge operation on entities, which modifies data structure by combining multiple entities into one while 'preserving all relationships'. This is a reversible structural transformation that alters the graph data model.
Attacks that exploit this kind of access
Merge multiple entities into a single entity, preserving all relationships. It is categorised as a Write tool in the Lightrag MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Lightrag MCP server in PolicyLayer and add a rule for merge_entities: 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.
merge_entities is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the merge_entities 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 merge_entities. 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.
merge_entities 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 →