Skip to main content
Glama

chrome_use_snapshot

Capture the current page's accessibility tree with @ref handles to locate elements for clicking, filling, or typing.

Instructions

Accessibility tree of the current page with @ref handles, for locating elements to click/fill/type.

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "chrome_use_snapshot"
  }
}

Input Schema

NameRequiredDescriptionDefault
diffNoOnly changes since this session last read the same page with the same options (--diff). Without a valid baseline returns the full tree and a diffNote. Omit or false for a full snapshot. Valid deltas include only excerpt refs; unchanged deltas have empty refs while the live session map is kept.
urlsNoInclude link/image URLs (-u).
depthNoLimit tree depth (-d).
forceNoWith tabId: act on a tab this session does not own (the user's own, or another session's), in place and in the background (--tab <handle> --force). Only when the user asked you to act on that tab; the result has forced: true and the tab's title and url.
tabIdNoTarget tab id (e.g. t2), label, or targetId (--tab).
cursorNoInclude the current cursor position marker (-C).
filterNoRegex: keep only matching lines + ancestor context (-f).
compactNoRemove empty structural elements (-c).
sessionNoIsolated session name (chrome-use --session <name>). Each session gets its own tab group, browser tab and element refs. Omit it unless several agents share ONE MCP server — the session is otherwise derived per agent automatically. When you do share a server with sibling agents, pass a name unique to you on every call, or you will drive their tabs and invalidate their refs.
selectorNoScope the snapshot to a CSS selector (-s).
timeoutMsNoKill the underlying chrome-use process if it hasn't responded after this many ms (default 30000).
interactiveNoOnly interactive elements (-i).
includeHiddenNoKeep text a person can't see on the page (off-screen, transparent, clipped, zero-size). Left out by default and counted in hiddenText/a note, because pages hide instructions aimed at agents there (--include-hidden).

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changedv1.5.275
    • addedInput schema / properties / includeHidden
      Added value: +{
      +  "description": "Keep text a person can't see on the page (off-screen, transparent, clipped, zero-size). Left out by default and counted in hiddenText/a note, because pages hide instructions aimed at agents there (--include-hidden).",
      +  "type": "boolean"
      +}
  2. First observedv0.1.0

TDQS

A3.6/5.0
Behavior3/5

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

With no annotations, the description must carry the behavioral burden; it discloses the return shape (accessibility tree + @ref handles), which implicitly frames this as a read operation. However, it says nothing about ref invalidation after navigation, that the snapshot reflects live page state, or how the result interacts with subsequent click/fill calls — the behaviors that matter most for this tool.

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

Conciseness5/5

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

A single front-loaded sentence with zero waste: resource, returned artifact, and intended follow-on actions in one line.

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?

For a read-only snapshot tool with fully documented params and no output schema, the description covers what it produces and why you'd call it. It is thin on ref lifetime and on how the snapshot relates to chrome_use_read, which would fully complete the picture.

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

Parameters3/5

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

Schema description coverage is 100% and each of the 13 parameters has a substantial inline description, so the schema already does the heavy lifting. The description adds nothing about diff/urls/depth/interactive/filter behavior, so the baseline 3 applies.

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?

States a specific verb+resource+output: the accessibility tree of the current page with @ref handles, and the intent (locating elements to click/fill/type). This clearly separates it from chrome_use_open or chrome_use_click, though it never names the most confusable sibling (chrome_use_read) to distinguish snapshot structure from raw text extraction.

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

Usage Guidelines3/5

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

The phrase 'for locating elements to click/fill/type' implies this is the discovery step that precedes interaction tools, but there is no explicit when/when-not guidance and no mention of the alternative chrome_use_read for text-only extraction. Usage is inferable, not stated.

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