web_eval
Runs arbitrary JavaScript in a web session and returns the last expression's value. POWER-USER tool, OFF BY DEFAULT (a page could feed malicious code) — enable "Allow web_eval (advanced)" in Local MCP settings first. Prefer web_find / web_read / web_extract for normal use.
This record as markdown: /tools/com-local-mcp-local-mcp/web-eval.md
What web_eval does on Local
AI agents invoke web_eval to trigger actions in Local. 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.
| Parameter | Type | Required | Description |
|---|---|---|---|
js | string | Yes | JavaScript to evaluate on the page; the last expression's value is returned. |
session | string | — | Session name (default 'default'). |
Parameters from the server's own tool schema.
Why web_eval is rated High
This tool executes arbitrary JavaScript in a browser context, which can perform virtually any action: exfiltrate data, modify DOM, make network requests, access credentials, and more. The description explicitly warns about malicious code injection risk and notes it is off by default due to danger. This is a classic arbitrary code execution vector with critical blast radius.
From the tool's definition Runs arbitrary JavaScript in a web session and returns the last expression's value. POWER-USER tool, OFF BY DEFAULT (a page could feed malicious code)
Risk signalsAccepts freeform code/query input (js)
Attacks that exploit this kind of access
The rule that runs web_eval safely
PolicyLayer is an MCP gateway: it sits between your AI agents and Local, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For web_eval, this is the rule to start with:
web_eval 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.
The button opens the PolicyLayer dashboard: create your workspace, connect Local, apply this rule, and every web_eval call is checked against it from then on.
Questions about web_eval
Runs arbitrary JavaScript in a web session and returns the last expression's value. POWER-USER tool, OFF BY DEFAULT (a page could feed malicious code) — enable "Allow web_eval (advanced)" in Local MCP settings first. Prefer web_find / web_read / web_extract for normal use. It is categorised as a Execute tool in the Local MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
web_eval accepts 2 parameters: js, session. Required: js. The full parameter table on this page comes from the server's own tool schema.
Register the Local MCP server in PolicyLayer and add a rule for web_eval: 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 Local. Nothing to install.
web_eval 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 web_eval 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 web_eval. 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.
web_eval is provided by the Local MCP server (local-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on Local, and thousands of servers like it.
Across the catalogue