AI agents use save_flow_model to create or update resources in Nocobase — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Nocobase environment.
The tool creates or updates flow models, which are data structures within NocoBase. While reversible (typical for update/create operations), it has a high severity due to the potential impact of modifying workflow/automation configurations that could affect business logic and processes. An AI agent could misuse this to alter critical flow logic without proper authorization.
From the tool's definition The tool 'save_flow_model' is described as allowing to 'Create or update a flowPage block/model' — these are explicit write operations that modify data reversibly.
Attacks that exploit this kind of access
Create or update a flowPage block/model. If. It is categorised as a Write tool in the Nocobase MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Nocobase MCP server in PolicyLayer and add a rule for save_flow_model: 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 Nocobase. Nothing to install.
save_flow_model 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 save_flow_model 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 save_flow_model. 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.
save_flow_model is provided by the Nocobase MCP server (puguhsudarma/nocobase-mcp-server). 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 →