AI agents invoke key_press to trigger actions in Yandex Browser MCP Server. 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.
Key press simulation triggers external operations in the browser whose effects depend on the arguments (e.g., pressing Enter could submit forms, Delete could remove content, Ctrl+W could close tabs). This is an Execute-level action with high blast radius as an AI agent could trigger destructive or unintended browser behaviors by pressing arbitrary key combinations.
From the tool's definition 'Нажать клавишу или комбинацию клавиш' (Press a key or key combination) — triggers browser/UI actions depending on which keys are pressed
Documented attack patterns abuse exactly the kind of access key_press gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Yandex Browser MCP Server, and nothing reaches the server without passing your rules. This is the rule we recommend for key_press:
{
"version": "1",
"default": "deny",
"tools": {
"key_press": {
"limits": [
{
"counter": "key_press_rate",
"window": "minute",
"max": 10,
"scope": "grant"
}
]
}
}
}key_press 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.
Нажать клавишу или комбинацию клавиш. It is categorised as a Execute tool in the Yandex Browser MCP Server MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Yandex Browser MCP Server MCP server in PolicyLayer and add a rule for key_press: 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 Yandex Browser MCP Server. Nothing to install.
key_press 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 key_press 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 key_press. 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.
key_press is provided by the Yandex Browser MCP Server MCP server (t1trit/yandex-browser-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Yandex Browser MCP Server, 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.
18 Yandex Browser MCP Server tools catalogued and risk-classified — across an index of 43,000+ MCP servers.