마지막 작업을 취소합니다.
AI agents use history_undo to create or update resources in SVG Canvas MCP — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your SVG Canvas MCP environment.
Undo operations are Write-category because they modify the current state of the document by reverting prior changes. Unlike Destructive operations, the undone changes are not permanently deleted—they can typically be restored via redo. Severity is medium because misuse could discard user work, but the effect is reversible. Context as part of a graphics editor reinforces this classification.
From the tool's definition Tool name 'history_undo' and description '마지막 작업을 취소합니다' (undo the last action) indicate reversal of prior modifications to SVG graphics. This modifies the canvas state by rolling back changes.
Attacks that exploit this kind of access
마지막 작업을 취소합니다. It is categorised as a Write tool in the SVG Canvas MCP MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the SVG Canvas MCP server in PolicyLayer and add a rule for history_undo: 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 SVG Canvas MCP. Nothing to install.
history_undo 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 history_undo 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 history_undo. 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.
history_undo is provided by the SVG Canvas MCP server (kim62210/svg-canvas-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 →