# 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.

Agent View of the PolicyLayer registry record for `web_eval`. HTML page: https://policylayer.com/tools/com-local-mcp-local-mcp/web-eval

## Facts

- Tool: `web_eval`
- Server: Local (`local-mcp`) — https://policylayer.com/tools/com-local-mcp-local-mcp.md
- Install: `npx -y local-mcp`
- Homepage: https://github.com/lanchuske/local-mcp
- Risk category: Execute (High risk)
- Registry record: grade F, identity unverified
- Server auth posture: open
- Server CORS policy: *
- Server rate-limited: no
- Parameters: 2 (1 required)
- Recommended policy verdict: Rate-limited

## Parameters

| Parameter | Type | Required | Description |
| --- | --- | --- | --- |
| `js` | string | yes | JavaScript to evaluate on the page; the last expression's value is returned. |
| `session` | string | no | Session name (default 'default'). |

Parameters from the server's own tool schema.

## Example call (MCP tools/call, JSON-RPC 2.0)

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "web_eval",
    "arguments": {
      "js": "<js>"
    }
  }
}
```

## 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 own 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 signals: Accepts freeform code/query input (js)

## Use case

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.

## Recommended policy (PolicyLayer)

Verdict: **Rate-limited**. Enforced by the PolicyLayer MCP gateway (https://policylayer.com/mcp-gateway) before a call reaches Local:

```json
{
  "version": "1",
  "default": "deny",
  "tools": {
    "web_eval": {
      "limits": [
        {
          "counter": "web_eval_rate",
          "window": "minute",
          "max": 10,
          "scope": "grant"
        }
      ]
    }
  }
}
```

## Other tools on Local (233)

- `delete_calendar_event` — Destructive — https://policylayer.com/tools/com-local-mcp-local-mcp/delete-calendar-event.md
- `delete_reminder` — Destructive — https://policylayer.com/tools/com-local-mcp-local-mcp/delete-reminder.md
- `delete_reminder_list` — Destructive — https://policylayer.com/tools/com-local-mcp-local-mcp/delete-reminder-list.md
- `m365_delete_event` — Destructive — https://policylayer.com/tools/com-local-mcp-local-mcp/m365-delete-event.md
- `onedrive_delete_file` — Destructive — https://policylayer.com/tools/com-local-mcp-local-mcp/onedrive-delete-file.md
- `recipe_delete` — Destructive — https://policylayer.com/tools/com-local-mcp-local-mcp/recipe-delete.md
- `chrome_click` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/chrome-click.md
- `chrome_evaluate_js` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/chrome-evaluate-js.md
- `chrome_fill_form` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/chrome-fill-form.md
- `chrome_go_back` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/chrome-go-back.md
- `chrome_navigate` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/chrome-navigate.md
- `chrome_type` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/chrome-type.md
- `chrome_wait_for` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/chrome-wait-for.md
- `nordvpn_diagnose` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/nordvpn-diagnose.md
- `notion_open_page` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/notion-open-page.md
- `recipe_run` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/recipe-run.md
- `run_diagnostics` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/run-diagnostics.md
- `run_terminal_command` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/run-terminal-command.md
- `safari_click` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/safari-click.md
- `safari_evaluate_js` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/safari-evaluate-js.md
- `safari_fill_form` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/safari-fill-form.md
- `safari_go_back` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/safari-go-back.md
- `safari_navigate` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/safari-navigate.md
- `safari_query_selector_all` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/safari-query-selector-all.md
- `safari_type` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/safari-type.md
- `safari_wait_for` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/safari-wait-for.md
- `screen_record_start` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/screen-record-start.md
- `screen_record_stop` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/screen-record-stop.md
- `setup_install` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/setup-install.md
- `ui_click` — Execute — https://policylayer.com/tools/com-local-mcp-local-mcp/ui-click.md
- …and 203 more: https://policylayer.com/tools/com-local-mcp-local-mcp.md

## For agents

This record is a snapshot. Live verdicts and the full registry:

- Check every server in your MCP config at once: `npx -y policylayer stack`
- Vet a server before you add it: install the mcp-precheck skill — `npx skills add https://policylayer.com` (skill text: https://policylayer.com/skill.md)
- Query the registry over MCP: endpoint `https://api.policylayer.com/mcp` — tools `check_mcp_server`, `check_mcp_stack`, `check_tool`, `search_registry`, `get_change_events`

---

Source: the PolicyLayer MCP registry — one continuously verified record per MCP server. Full record: https://policylayer.com/registry?q=com-local-mcp-local-mcp · API: https://policylayer.com/registry/api · Policy library: https://policylayer.com/policies/com-local-mcp-local-mcp