AI agents use crop_image to create or update resources in Graphics — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Graphics environment.
Cropping an image modifies it by removing portions of the image data, creating a new version. This is reversible (the original can be restored or the operation undone), so it falls under Write rather than Destructive. The severity is medium because widespread image modification could affect many files, though the blast radius is limited to image data without affecting systems or financial operations. Confidence is 0.
From the tool's definition Tool name is 'crop_image' with no description provided. Sibling tools on the graphics-mcp server (apply_filter, convert_image, create_thumbnail, flip_image, resize_image, rotate_image) are all image manipulation operations that modify images.
Attacks that exploit this kind of access
crop_image. It is categorised as a Write tool in the Graphics MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Graphics MCP server in PolicyLayer and add a rule for crop_image: 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 Graphics. Nothing to install.
crop_image 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 crop_image 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 crop_image. 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.
crop_image is provided by the Graphics MCP server (lesleslie/graphics-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 →