screen.observe
Read the current screen: UIA targets (numbered ids + native coords) and a text summary. Does not move mouse/keyboard. Default image=false (no JPEG) for speed; set image=true only when you must judge pixels visually (then max_width≈960, quality≈60). If changed=false, JPEG is omitted even when requ...
This record as markdown: /tools/com-glasswarp-mcp-server/screen.observe.md
What screen.observe does on Glasswarp
AI agents call screen.observe to retrieve information from Glasswarp without modifying anything. It is typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
| Parameter | Type | Required | Description |
|---|---|---|---|
mark | boolean | — | Overlay numbered targets on JPEG when image=true (default true; ignored if image=false) |
image | boolean | — | Include JPEG (default false). True only to visually read/judge the screen. |
quality | integer | — | JPEG quality 40–100 when image=true (default 60) |
max_width | integer | — | JPEG max width when image=true (default 960). Clicks always use native coords, not JPEG pixels. |
session_id | string | Yes | Active session id from session.start |
Parameters from the server's own tool schema.
Why screen.observe is rated Low
Purely observes screen state with no side effects or modifications.
From the tool's definition Read the current screen... Does not move mouse/keyboard
Risk signalsBulk/mass operation — affects multiple targets
Attacks that exploit this kind of access
The rule that runs screen.observe safely
PolicyLayer is an MCP gateway: it sits between your AI agents and Glasswarp, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For screen.observe, this is the rule to start with:
screen.observe is read-only, so it stays allowed. Everything else on the server is denied unless you say otherwise.
The button opens the PolicyLayer dashboard: create your workspace, connect Glasswarp, apply this rule, and every screen.observe call is checked against it from then on.
Questions about screen.observe
Read the current screen: UIA targets (numbered ids + native coords) and a text summary. Does not move mouse/keyboard. Default image=false (no JPEG) for speed; set image=true only when you must judge pixels visually (then max_width≈960, quality≈60). If changed=false, JPEG is omitted even when requested — do not re-analyze; wait or act differently. If dirty is null, assume changed. Prefer input.send_actions for multi-step UI; observe after meaningful steps, not after every click. Target ids are valid only until the next UI change. It is categorised as a Read tool in the Glasswarp MCP Server, which means it retrieves data without modifying state.
screen.observe accepts 5 parameters: mark, image, quality, max_width, session_id. Required: session_id. The full parameter table on this page comes from the server's own tool schema.
Register the Glasswarp MCP server in PolicyLayer and add a rule for screen.observe: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Glasswarp. Nothing to install.
screen.observe is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the screen.observe rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for screen.observe. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
screen.observe is provided by the Glasswarp MCP server (https://mcp.glasswarp.com/mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on Glasswarp, and thousands of servers like it.
This server
Across the catalogue