AI agents use crop to create or update resources in Ffmpeg — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Ffmpeg environment.
The crop operation transforms video data in a reversible manner—the original file remains unchanged and the output is a modified version. This fits the Write category (modifies data reversibly) rather than Destructive (irreversible deletion).
From the tool's definition Tool 'crop' modifies video content by changing dimensions and region ('Crop video to a specific region. Specify width, height, and position offset'). This creates a new processed output without permanently destroying the original source file.
Attacks that exploit this kind of access
Crop video to a specific region. Specify width, height, and position offset. It is categorised as a Write tool in the Ffmpeg MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Ffmpeg MCP server in PolicyLayer and add a rule for crop: 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 Ffmpeg. Nothing to install.
crop 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 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. 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 is provided by the Ffmpeg MCP server (yuan66-hub/ffmpeg-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 →