AI agents call get_canvas_pixels to retrieve information from Paint MCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool queries and retrieves pixel data from the drawing canvas without creating, modifying, deleting, or executing any external operations. It is purely informational, similar to a 'get' or 'fetch' operation. The return of pixel data as a JSON array poses minimal security risk—an LLM misusing this tool would only read data already present on the canvas, with no ability to alter it or trigger external effects.
From the tool's definition Tool name 'get_canvas_pixels' and description 'Return RGB pixel data from the canvas as a JSON 2D array' indicate retrieval of existing canvas state with no side effects or modifications.
Attacks that exploit this kind of access
Return RGB pixel data from the canvas as a JSON 2D array of [r,g,b] values (row-major). It is categorised as a Read tool in the Paint MCP MCP Server, which means it retrieves data without modifying state.
Register the Paint MCP server in PolicyLayer and add a rule for get_canvas_pixels: 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 Paint MCP. Nothing to install.
get_canvas_pixels is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the get_canvas_pixels 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 get_canvas_pixels. 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.
get_canvas_pixels is provided by the Paint MCP server (joeyballentine/paint-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 →