AI agents use create_record to create or update resources in Devon — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Devon environment.
Creating records modifies the DEVONthink database by adding new data. This falls squarely into the Write category as it creates data reversibly. Severity is medium because an AI agent could spam the database with unwanted records, but the impact is not directly financial, destructive (irreversible), or operationally critical compared to executing arbitrary commands.
From the tool's definition Tool name 'create_record' and description 'Create a new record in DEVONthink' indicate creating new data in the system. This is reversible (records can be deleted via 'delete_record' sibling tool).
Attacks that exploit this kind of access
Create a new record in DEVONthink. It is categorised as a Write tool in the Devon MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Devon MCP server in PolicyLayer and add a rule for create_record: 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 Devon. Nothing to install.
create_record 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 create_record 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 create_record. 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.
create_record is provided by the Devon MCP server (mnott/devon). 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 →