AI Commands Reference

Command Reference

Complete reference for all AI commands supported by Comet Browser. Commands are issued as structured JSON for precise execution, though natural language is also understood.

Risk Levels:

Low
Medium
High

CREATE_PDF_JSON

JSON Format

Low Risk

Generate a structured, multi-page PDF document with advanced formatting options.

Parameters (2 required)

titletemplatecontentbranding+1 more
Instant

GENERATE_PDF

Deprecated

JSON Format

Low Risk

Legacy PDF generation command. Prefer CREATE_PDF_JSON for new implementations.

Parameters (1 required)

contentfilename
Instant

SHELL_COMMAND

JSON Format

High Risk

Execute terminal commands on the host operating system. All shell commands require explicit user approval.

Parameters (1 required)

commandshell
Requires Approval

NAVIGATE

JSON Format

Low Risk

Navigate to a specific URL or perform a web search.

Parameters (1 required)

urlnewTabbackground
Instant

CLICK_ELEMENT

JSON Format

Medium Risk

Click on a visible DOM element using visual intelligence (OCR + coordinate mapping).

Parameters (1 required)

textmatchTypeindex
Requires Approval

FIND_AND_CLICK

JSON Format

Medium Risk

Find text on the current page and click nearby elements. Uses OCR for text detection.

Parameters (1 required)

findTextclickText
Requires Approval

FILL_FORM

JSON Format

Medium Risk

Fill in form fields identified by label, placeholder, or name.

Parameters (2 required)

fieldsfields[].labelfields[].namefields[].value
Requires Approval

SCROLL

JSON Format

Low Risk

Scroll the page by a specified amount or to a specific element.

Parameters (1 required)

directionamount
Instant

SET_VOLUME

JSON Format

Low Risk

Adjust the system volume level.

Parameters (1 required)

levelrelative
Instant

SET_BRIGHTNESS

JSON Format

Low Risk

Adjust the display brightness level.

Parameters (1 required)

level
Instant

OPEN_APP

JSON Format

Medium Risk

Launch an application on the host system.

Parameters (1 required)

appurl
Requires Approval

OCR_SCREEN

JSON Format

Low Risk

Capture the screen and extract text using OCR. Useful for reading content from non-web sources.

Parameters (0 required)

region
Instant

OCR_COORDINATES

JSON Format

Low Risk

Get screen coordinates for a specific text element. Used for precise clicking.

Parameters (1 required)

textthreshold
Instant

CLICK_APP_ELEMENT

JSON Format

High Risk

Click on an element in an external (non-browser) application using OCR coordinates.

Parameters (2 required)

apptextxy
Requires Approval

SCHEDULE_TASK

JSON Format

Low Risk

Schedule a task to run automatically at specified times. Tasks run even when browser is closed.

Parameters (3 required)

taskTypeschedulepromptmodel+1 more
Instant

Quick Command Reference

CREATE_PDF_JSON
GENERATE_PDF
SHELL_COMMAND
NAVIGATE
CLICK_ELEMENT
FIND_AND_CLICK
FILL_FORM
SCROLL
SET_VOLUME

Understand the Security Model

Learn how Comet AI protects your system with visual sandboxing, syntactic firewalls, and human-in-the-loop authorization.

Edit on GitHub