Skip to main content
Glama
scrape-mcp
by guanyuyan

web_inspect

Scan pages read-only to return clickable and input elements, forms, and pagination candidates with stable CSS selector refs for input-click-wait automation.

Instructions

只读扫描页面,列出可点击/可输入元素、表单与分页候选。

返回的每个元素都带 ref(稳定 CSS 选择器),可直接放进 web_fetch / web_screenshot 的 actions[].selector 执行"输入 → 点击 → 等待", 不需要模型猜选择器。只读:不执行任意 JS,不返回整页 HTML。

Args: url: 目标网页完整 URL。 max_elements: 最多返回多少个元素;不传取配置 inspect_max_elements(默认 120)。 session: 已登录会话名,渲染时带上该会话的 cookie。 deadline_ms: 整次调用的总预算(毫秒),不传表示不限。 proxy: 本次请求使用的代理;不传则按代理池分配。 region: 地区定向(如 us / hk / jp),只在走代理池时生效。

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "url": "<url>"
    },
    "name": "web_inspect"
  }
}

Input Schema

NameRequiredDescriptionDefault
urlYes
proxyNo
regionNo
sessionNodefault
deadline_msNo
max_elementsNo

Output Schema

NameRequiredDescriptionDefault
resultYes

Schema Changelog

Changes observed during successful MCP inspections.

  1. Addedv0.35.0

TDQS

A4.2/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the safety burden and does so well: read-only, no arbitrary JS execution, no full-page HTML returned, plus a hard total time budget via deadline_ms and proxy-pool vs. direct-request distinction. Still undisclosed are timeout/error behavior when deadline_ms is exceeded, any rate or cost limits, and what happens on login-walled pages even with a session.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded: purpose and the ref payoff land in the first two sentences, then a clean Args list, one line per parameter. Some redundancy ('只读' repeated, config-name spelling out) but no sentence is wasted for a six-parameter tool.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

An output schema exists, so return shape need not be explained, and the description still clarifies the one field that matters (`ref`). All six parameters are documented despite 0% schema coverage. The remaining gap is sibling routing — which tool to reach for when element discovery is not what is wanted.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 0%, so the Args block is the only documentation and it covers all six parameters with semantics the schema lacks: max_elements falls back to inspect_max_elements (default 120), session injects that session's cookies at render time, deadline_ms is a whole-call budget with unlimited as default, and region only applies when routing through the proxy pool.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

Opens with a specific verb+resource: a read-only page scan that enumerates clickable/typable elements, forms and pagination candidates. It also states the distinguishing output (stable CSS `ref` per element), which separates it from content-oriented siblings. It does not explicitly contrast with web_extract or web_map, so an agent must infer that those are for content rather than element discovery.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives concrete workflow guidance: feed the returned `ref` into web_fetch / web_screenshot `actions[].selector` to perform input → click → wait without guessing selectors. That is real when-to-use context. It stops short of exclusions — nothing says when to prefer web_extract, web_map or raw web_fetch instead.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.