visual_check · Perceptdot MCP: Risk & Policy

WHAT IT DOES

What visual_check does on Perceptdot

AI agents invoke visual_check to trigger actions in Perceptdot. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call — builds kicked off, notifications sent, workflows started.

WHY IT NEEDS A POLICY

Why visual_check needs a policy

The tool performs an active external operation (fetching and screenshotting a URL) and then executes AI-based analysis. This goes beyond passive reading — it drives a browser action against an external target and processes the result, fitting the Execute category. Misuse could involve screenshotting sensitive internal URLs or consuming resources against arbitrary targets.

From the tool's definition 'Screenshot a URL and analyze it' — triggers an external browser/rendering operation against an arbitrary URL and runs AI analysis on the result

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

visual_check Rate-limited

visual_check stays usable, but rate-capped — a runaway agent can't fire it dozens of times a minute. Everything else on the server is denied unless you say otherwise.

View as policy code

policy.json

{
  "version": "1",
  "default": "deny",
  "tools": {
    "visual_check": {
      "limits": [
        {
          "counter": "visual_check_rate",
          "window": "minute",
          "max": 10,
          "scope": "grant"
        }
      ]
    }
  }
}

RATE-LIMIT THIS TOOL →

Instant setup, no code required.

PolicyLayer is an MCP gateway: every tool call is checked against your rule before it reaches Perceptdot, so you can run agents against it safely.

FAQ

Questions about visual_check

What does the visual_check tool do? +

Screenshot a URL and analyze it for visual bugs using AI. It is categorised as a Execute tool in the Perceptdot MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

How do I enforce a policy on visual_check? +

Register the Perceptdot MCP server in PolicyLayer and add a rule for visual_check: 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 Perceptdot. Nothing to install.

What risk level is visual_check? +

visual_check is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.

Can I rate-limit visual_check? +

Yes. Add a rate_limit block to the visual_check 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.

How do I block visual_check completely? +

Set action: deny in the PolicyLayer policy for visual_check. 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.

What MCP server provides visual_check? +

visual_check is provided by the Perceptdot MCP server (perceptdot/percept). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.