给网页拍一张截图（境内真 Chromium 渲染），PNG 落文件中转站并返 15 分钟下载 URL。
比 browser_run 简单直接：一个 URL 进、一张图出；固定 1 credit，失败不计费。
| Name | Required | Description | Default |
|---|---|---|---|
| url | Yes | 要截图的网页 URL（http/https，仅限公开网页） | |
| wait_ms | No | 截图前等待毫秒（等 JS 渲染完），默认 1500 | |
| full_page | No | 整页长图（false=只截首屏视口） |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of disclosure. It transparently states the rendering environment (real Chromium), output format (PNG), temporary download URL (15 minutes), and cost policy (fixed 1 credit, free on failure). This adds meaningful behavioral context beyond the schema.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is two sentences: the first states the core function and output, the second adds a sibling comparison and cost detail. Every sentence serves a purpose, with no redundant content or clutter.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a simple tool with no output schema, this description explains the output format, URL expiration, pricing, and failure behavior, covering key context. It could mention browser_pdf as another alternative, but the tool is simple enough that the description feels sufficiently complete.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema provides 100% parameter coverage with descriptions for url, wait_ms, and full_page. The description adds only the generic 'one URL in' concept and does not deepen understanding of the parameters beyond what the schema already offers, so the baseline of 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool takes a screenshot of a webpage with a specific verb ('拍一张截图') and resource ('网页'). It explicitly differentiates from browser_run by noting it's simpler and returns one image from one URL, making the purpose unmistakable.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives explicit usage context by comparing to browser_run ('比 browser_run 简单直接') and describing the one-in-one-out flow. It implies this tool is preferred for straightforward screenshots, though it does not mention browser_pdf as an alternative, leaving some room for interpretation.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Add one secure layer between your agents and this server.
Each tool has a clearly distinct purpose: PDF for document capture, screenshot for image capture, and run for interactive extraction with data output. Although run also returns a screenshot, its primary role as an action executor is unambiguous from the description.
All tools share the consistent 'browser_' prefix, but the suffixes mix nouns (pdf, screenshot) and a verb (run). This is a minor deviation from a pure verb_noun pattern, yet the overall naming scheme is predictable and readable.
With just three tools, the set is tightly focused on the core browser operations: print, capture, and interact. This is well within the ideal 3-15 range and each tool earns its place.
The tool surface covers the primary output methods (PDF, screenshot, data extraction) and the run tool is generic enough to handle various interaction scripts. A minor gap might be the lack of a direct HTML/raw content tool, but agents can work around this using run.