AI agents invoke browser_evaluate to trigger actions in Playwright. 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.
In Playwright, 'evaluate' refers to executing arbitrary JavaScript in the browser page context. This allows running arbitrary code with full access to the DOM, network, storage, and potentially sensitive data. The empty description lowers confidence slightly, but the tool name combined with the Playwright server context makes this classification highly probable.
From the tool's definition Tool name 'browser_evaluate' strongly implies execution of JavaScript or expressions in a browser context, consistent with Playwright's page.evaluate() API which runs arbitrary code in the browser.
Documented attack patterns abuse exactly the kind of access browser_evaluate gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Playwright, and nothing reaches the server without passing your rules. This is the rule we recommend for browser_evaluate:
{
"version": "1",
"default": "deny",
"tools": {
"browser_evaluate": {
"limits": [
{
"counter": "browser_evaluate_rate",
"window": "minute",
"max": 10,
"scope": "grant"
}
]
}
}
}browser_evaluate 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.
Free to start. No card required.
browser_evaluate. It is categorised as a Execute tool in the Playwright MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Playwright MCP server in PolicyLayer and add a rule for browser_evaluate: 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 Playwright. Nothing to install.
browser_evaluate is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the browser_evaluate 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 browser_evaluate. 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.
browser_evaluate is provided by the Playwright MCP server (@playwright/mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Playwright, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
68 Playwright tools catalogued and risk-classified — across an index of 43,000+ MCP servers.