Skip to main content
Glama

ghost_see

Describe a window's UI as structured elements (name, role, rect) by matching a title substring or using the foreground window. Read visible text, detect changes, or retrieve selected text via modes.

Instructions

Describe a window's UI as elements {name, role, rect}. Target = window= (title substring; becomes the session anchor) or the current anchor; with neither, the human's foreground window and the response says so (target.source). Never focuses or raises anything. mode=full (default whenever a window is targeted: scoped walk, works on covered windows and hidden-desktop apps), mode=fast (foreground-only quick path, used only when nothing is targeted), mode=text (READ the visible text - the cheapest way to read content), mode=delta (changed elements since since_seq), mode=marks (DEBUG Set-of-Marks image), mode=selection (selected text of a name/role element). Elements capped at 150 (limit); text at 20000 chars.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
modeNofull=scoped tree (default with a target), fast=foreground quick path, delta=changed only, text=readable text, marks=SoM debug image, selection=selected text of name/role element
nameNoElement name (mode=selection)
roleNoElement role (mode=selection)
limitNoMax elements (default 150) or chars for mode=text (default 20000); 0 = unlimited elements
windowNoTitle substring of the window to describe. Anchors it for later calls. 'foreground' = the human's current window.
since_seqNoPrior snapshot seq for delta mode

Schema Changelog

Changes observed during successful MCP inspections. Dates show when Glama detected each change.

  1. First observedv0.23.4

TDQS

A4.7/5.0
Behavior5/5

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

No annotations are provided, so the description carries the full burden—and it delivers. It discloses that the tool 'Never focuses or raises anything', how session anchoring works, what happens when neither window nor anchor is provided, coverage of hidden-desktop windows, and hard caps on elements and text.

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?

The description is dense but every clause earns its place: target resolution, no-focus guarantee, per-mode semantics, and limits. It is front-loaded with the core purpose and preserves a clear structure through semicolon-separated mode explanations, making it efficient for a tool with six parameters and multiple modes.

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

Completeness5/5

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

With no output schema and no annotations, the description must cover return shape and side-effect behavior itself. It states the element shape, target source reporting, mode-specific outputs like 'SoM debug image' and 'selected text', plus caps for elements and text. This is sufficient for an agent to invoke the tool correctly across all modes.

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

Parameters4/5

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

Schema coverage is 100%, so the baseline is 3. The description adds meaningful extras beyond the schema: window= 'becomes the session anchor', mode=full is the default when a window is targeted, fast is only for the foreground quick path, and text is positioned as the cheapest read path. These details improve parameter selection beyond the schema alone.

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

Purpose5/5

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

The opening sentence states a specific verb and resource: 'Describe a window's UI as elements {name, role, rect}.' This clearly sets it apart from sibling tab-focused or action-oriented tools, and the target/window semantics solidify what object the tool operates on.

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?

The description gives explicit guidance on when each mode applies, e.g., 'full (default whenever a window is targeted)', 'fast (foreground-only quick path, used only when nothing is targeted)', and 'text (READ the visible text - the cheapest way to read content)'. It stops short of explicitly contrasting this tool with sibling tools like ghost_find or ghost_tab_describe, but the mode-level usage guidance is strong.

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

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/NORTHTEKDevs/ghost'

If you have feedback or need assistance with the MCP directory API, please join our Discord server