Create a new knowledge base
AI agents use knowledge_create_dataset to create or update resources in Dify Management MCP — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Dify Management MCP environment.
Creating a knowledge base is a write operation that modifies the state of the system by adding new data structures, but it is reversible (the dataset can be deleted). It does not execute arbitrary code, delete irreversibly, move money, or merely read data.
From the tool's definition Tool name 'knowledge_create_dataset' and description 'Create a new knowledge base' indicate data creation. This is a reversible write operation that adds a new dataset/knowledge base to the Dify instance.
Attacks that exploit this kind of access
Create a new knowledge base. It is categorised as a Write tool in the Dify Management MCP MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Dify Management MCP server in PolicyLayer and add a rule for knowledge_create_dataset: 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 Dify Management MCP. Nothing to install.
knowledge_create_dataset 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 knowledge_create_dataset 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 knowledge_create_dataset. 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.
knowledge_create_dataset is provided by the Dify Management MCP server (lucasallmabi/mcp-dify). 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 →