Kernel Browser Computer Controls API
Control mouse, keyboard, and screen on the browser instance.
OpenAPI Specification
openapi: 3.1.0
info:
title: Kernel API Keys Browser Computer Controls API
description: Developer tools and cloud infrastructure for AI agents to use web browsers
version: 0.1.0
servers:
- url: https://api.onkernel.com
description: API Server
security:
- bearerAuth: []
tags:
- name: Browser Computer Controls
description: Control mouse, keyboard, and screen on the browser instance.
paths:
/browsers/{id}/computer/click_mouse:
post:
summary: Simulate a mouse click action on the browser instance
operationId: clickMouse
tags:
- Browser Computer Controls
security:
- bearerAuth: []
parameters:
- name: id
in: path
required: true
schema:
type: string
description: Browser session ID
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/ClickMouseRequest'
responses:
'200':
description: Mouse action performed
'400':
$ref: '#/components/responses/BadRequest'
'404':
$ref: '#/components/responses/NotFound'
'500':
$ref: '#/components/responses/InternalError'
x-codeSamples:
- lang: JavaScript
source: "import Kernel from '@onkernel/sdk';\n\nconst client = new Kernel({\n apiKey: process.env['KERNEL_API_KEY'], // This is the default and can be omitted\n});\n\nawait client.browsers.computer.clickMouse('id', { x: 0, y: 0 });"
- lang: Python
source: "import os\nfrom kernel import Kernel\n\nclient = Kernel(\n api_key=os.environ.get(\"KERNEL_API_KEY\"), # This is the default and can be omitted\n)\nclient.browsers.computer.click_mouse(\n id=\"id\",\n x=0,\n y=0,\n)"
- lang: Go
source: "package main\n\nimport (\n\t\"context\"\n\n\t\"github.com/kernel/kernel-go-sdk\"\n\t\"github.com/kernel/kernel-go-sdk/option\"\n)\n\nfunc main() {\n\tclient := kernel.NewClient(\n\t\toption.WithAPIKey(\"My API Key\"),\n\t)\n\terr := client.Browsers.Computer.ClickMouse(\n\t\tcontext.TODO(),\n\t\t\"id\",\n\t\tkernel.BrowserComputerClickMouseParams{\n\t\t\tX: 0,\n\t\t\tY: 0,\n\t\t},\n\t)\n\tif err != nil {\n\t\tpanic(err.Error())\n\t}\n}\n"
/browsers/{id}/computer/move_mouse:
post:
summary: Move the mouse cursor to the specified coordinates on the browser instance
operationId: moveMouse
tags:
- Browser Computer Controls
security:
- bearerAuth: []
parameters:
- name: id
in: path
required: true
schema:
type: string
description: Browser session ID
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/MoveMouseRequest'
responses:
'200':
description: Mouse cursor moved
'400':
$ref: '#/components/responses/BadRequest'
'404':
$ref: '#/components/responses/NotFound'
'500':
$ref: '#/components/responses/InternalError'
x-codeSamples:
- lang: JavaScript
source: "import Kernel from '@onkernel/sdk';\n\nconst client = new Kernel({\n apiKey: process.env['KERNEL_API_KEY'], // This is the default and can be omitted\n});\n\nawait client.browsers.computer.moveMouse('id', { x: 0, y: 0 });"
- lang: Python
source: "import os\nfrom kernel import Kernel\n\nclient = Kernel(\n api_key=os.environ.get(\"KERNEL_API_KEY\"), # This is the default and can be omitted\n)\nclient.browsers.computer.move_mouse(\n id=\"id\",\n x=0,\n y=0,\n)"
- lang: Go
source: "package main\n\nimport (\n\t\"context\"\n\n\t\"github.com/kernel/kernel-go-sdk\"\n\t\"github.com/kernel/kernel-go-sdk/option\"\n)\n\nfunc main() {\n\tclient := kernel.NewClient(\n\t\toption.WithAPIKey(\"My API Key\"),\n\t)\n\terr := client.Browsers.Computer.MoveMouse(\n\t\tcontext.TODO(),\n\t\t\"id\",\n\t\tkernel.BrowserComputerMoveMouseParams{\n\t\t\tX: 0,\n\t\t\tY: 0,\n\t\t},\n\t)\n\tif err != nil {\n\t\tpanic(err.Error())\n\t}\n}\n"
/browsers/{id}/computer/screenshot:
post:
summary: Capture a screenshot of the browser instance
operationId: takeScreenshot
tags:
- Browser Computer Controls
security:
- bearerAuth: []
parameters:
- name: id
in: path
required: true
schema:
type: string
description: Browser session ID
requestBody:
required: false
content:
application/json:
schema:
$ref: '#/components/schemas/ScreenshotRequest'
responses:
'200':
description: Screenshot image
content:
image/png:
schema:
type: string
format: binary
'400':
$ref: '#/components/responses/BadRequest'
'404':
$ref: '#/components/responses/NotFound'
'500':
$ref: '#/components/responses/InternalError'
x-codeSamples:
- lang: JavaScript
source: "import Kernel from '@onkernel/sdk';\n\nconst client = new Kernel({\n apiKey: process.env['KERNEL_API_KEY'], // This is the default and can be omitted\n});\n\nconst response = await client.browsers.computer.captureScreenshot('id');\n\nconsole.log(response);\n\nconst content = await response.blob();\nconsole.log(content);"
- lang: Python
source: "import os\nfrom kernel import Kernel\n\nclient = Kernel(\n api_key=os.environ.get(\"KERNEL_API_KEY\"), # This is the default and can be omitted\n)\nresponse = client.browsers.computer.capture_screenshot(\n id=\"id\",\n)\nprint(response)\ncontent = response.read()\nprint(content)"
- lang: Go
source: "package main\n\nimport (\n\t\"context\"\n\t\"fmt\"\n\n\t\"github.com/kernel/kernel-go-sdk\"\n\t\"github.com/kernel/kernel-go-sdk/option\"\n)\n\nfunc main() {\n\tclient := kernel.NewClient(\n\t\toption.WithAPIKey(\"My API Key\"),\n\t)\n\tresponse, err := client.Browsers.Computer.CaptureScreenshot(\n\t\tcontext.TODO(),\n\t\t\"id\",\n\t\tkernel.BrowserComputerCaptureScreenshotParams{},\n\t)\n\tif err != nil {\n\t\tpanic(err.Error())\n\t}\n\tfmt.Printf(\"%+v\\n\", response)\n}\n"
/browsers/{id}/computer/type:
post:
summary: Type text on the browser instance
operationId: typeText
tags:
- Browser Computer Controls
security:
- bearerAuth: []
parameters:
- name: id
in: path
required: true
schema:
type: string
description: Browser session ID
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/TypeTextRequest'
responses:
'200':
description: Text typed successfully
'400':
$ref: '#/components/responses/BadRequest'
'404':
$ref: '#/components/responses/NotFound'
'500':
$ref: '#/components/responses/InternalError'
x-codeSamples:
- lang: JavaScript
source: "import Kernel from '@onkernel/sdk';\n\nconst client = new Kernel({\n apiKey: process.env['KERNEL_API_KEY'], // This is the default and can be omitted\n});\n\nawait client.browsers.computer.typeText('id', { text: 'text' });"
- lang: Python
source: "import os\nfrom kernel import Kernel\n\nclient = Kernel(\n api_key=os.environ.get(\"KERNEL_API_KEY\"), # This is the default and can be omitted\n)\nclient.browsers.computer.type_text(\n id=\"id\",\n text=\"text\",\n)"
- lang: Go
source: "package main\n\nimport (\n\t\"context\"\n\n\t\"github.com/kernel/kernel-go-sdk\"\n\t\"github.com/kernel/kernel-go-sdk/option\"\n)\n\nfunc main() {\n\tclient := kernel.NewClient(\n\t\toption.WithAPIKey(\"My API Key\"),\n\t)\n\terr := client.Browsers.Computer.TypeText(\n\t\tcontext.TODO(),\n\t\t\"id\",\n\t\tkernel.BrowserComputerTypeTextParams{\n\t\t\tText: \"text\",\n\t\t},\n\t)\n\tif err != nil {\n\t\tpanic(err.Error())\n\t}\n}\n"
/browsers/{id}/computer/press_key:
post:
summary: Press one or more keys on the host computer
operationId: pressKey
tags:
- Browser Computer Controls
security:
- bearerAuth: []
parameters:
- name: id
in: path
required: true
schema:
type: string
description: Browser session ID
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/PressKeyRequest'
responses:
'200':
description: Keys pressed successfully
'400':
$ref: '#/components/responses/BadRequest'
'404':
$ref: '#/components/responses/NotFound'
'500':
$ref: '#/components/responses/InternalError'
x-codeSamples:
- lang: JavaScript
source: "import Kernel from '@onkernel/sdk';\n\nconst client = new Kernel({\n apiKey: process.env['KERNEL_API_KEY'], // This is the default and can be omitted\n});\n\nawait client.browsers.computer.pressKey('id', { keys: ['string'] });"
- lang: Python
source: "import os\nfrom kernel import Kernel\n\nclient = Kernel(\n api_key=os.environ.get(\"KERNEL_API_KEY\"), # This is the default and can be omitted\n)\nclient.browsers.computer.press_key(\n id=\"id\",\n keys=[\"string\"],\n)"
- lang: Go
source: "package main\n\nimport (\n\t\"context\"\n\n\t\"github.com/kernel/kernel-go-sdk\"\n\t\"github.com/kernel/kernel-go-sdk/option\"\n)\n\nfunc main() {\n\tclient := kernel.NewClient(\n\t\toption.WithAPIKey(\"My API Key\"),\n\t)\n\terr := client.Browsers.Computer.PressKey(\n\t\tcontext.TODO(),\n\t\t\"id\",\n\t\tkernel.BrowserComputerPressKeyParams{\n\t\t\tKeys: []string{\"string\"},\n\t\t},\n\t)\n\tif err != nil {\n\t\tpanic(err.Error())\n\t}\n}\n"
/browsers/{id}/computer/scroll:
post:
summary: Scroll the mouse wheel at a position on the host computer
operationId: scroll
tags:
- Browser Computer Controls
security:
- bearerAuth: []
parameters:
- name: id
in: path
required: true
schema:
type: string
description: Browser session ID
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/ScrollRequest'
responses:
'200':
description: Scroll performed
'400':
$ref: '#/components/responses/BadRequest'
'404':
$ref: '#/components/responses/NotFound'
'500':
$ref: '#/components/responses/InternalError'
x-codeSamples:
- lang: JavaScript
source: "import Kernel from '@onkernel/sdk';\n\nconst client = new Kernel({\n apiKey: process.env['KERNEL_API_KEY'], // This is the default and can be omitted\n});\n\nawait client.browsers.computer.scroll('id', { x: 0, y: 0 });"
- lang: Python
source: "import os\nfrom kernel import Kernel\n\nclient = Kernel(\n api_key=os.environ.get(\"KERNEL_API_KEY\"), # This is the default and can be omitted\n)\nclient.browsers.computer.scroll(\n id=\"id\",\n x=0,\n y=0,\n)"
- lang: Go
source: "package main\n\nimport (\n\t\"context\"\n\n\t\"github.com/kernel/kernel-go-sdk\"\n\t\"github.com/kernel/kernel-go-sdk/option\"\n)\n\nfunc main() {\n\tclient := kernel.NewClient(\n\t\toption.WithAPIKey(\"My API Key\"),\n\t)\n\terr := client.Browsers.Computer.Scroll(\n\t\tcontext.TODO(),\n\t\t\"id\",\n\t\tkernel.BrowserComputerScrollParams{\n\t\t\tX: 0,\n\t\t\tY: 0,\n\t\t},\n\t)\n\tif err != nil {\n\t\tpanic(err.Error())\n\t}\n}\n"
/browsers/{id}/computer/drag_mouse:
post:
summary: Drag the mouse along a path
operationId: dragMouse
tags:
- Browser Computer Controls
security:
- bearerAuth: []
parameters:
- name: id
in: path
required: true
schema:
type: string
description: Browser session ID
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/DragMouseRequest'
responses:
'200':
description: Drag performed
'400':
$ref: '#/components/responses/BadRequest'
'404':
$ref: '#/components/responses/NotFound'
'500':
$ref: '#/components/responses/InternalError'
x-codeSamples:
- lang: JavaScript
source: "import Kernel from '@onkernel/sdk';\n\nconst client = new Kernel({\n apiKey: process.env['KERNEL_API_KEY'], // This is the default and can be omitted\n});\n\nawait client.browsers.computer.dragMouse('id', {\n path: [\n [0, 0],\n [0, 0],\n ],\n});"
- lang: Python
source: "import os\nfrom kernel import Kernel\n\nclient = Kernel(\n api_key=os.environ.get(\"KERNEL_API_KEY\"), # This is the default and can be omitted\n)\nclient.browsers.computer.drag_mouse(\n id=\"id\",\n path=[[0, 0], [0, 0]],\n)"
- lang: Go
source: "package main\n\nimport (\n\t\"context\"\n\n\t\"github.com/kernel/kernel-go-sdk\"\n\t\"github.com/kernel/kernel-go-sdk/option\"\n)\n\nfunc main() {\n\tclient := kernel.NewClient(\n\t\toption.WithAPIKey(\"My API Key\"),\n\t)\n\terr := client.Browsers.Computer.DragMouse(\n\t\tcontext.TODO(),\n\t\t\"id\",\n\t\tkernel.BrowserComputerDragMouseParams{\n\t\t\tPath: [][]int64{{0, 0}, {0, 0}},\n\t\t},\n\t)\n\tif err != nil {\n\t\tpanic(err.Error())\n\t}\n}\n"
/browsers/{id}/computer/cursor:
post:
summary: Set cursor visibility
operationId: setCursor
tags:
- Browser Computer Controls
security:
- bearerAuth: []
parameters:
- name: id
in: path
required: true
schema:
type: string
description: Browser session ID
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/SetCursorRequest'
responses:
'200':
description: Cursor visibility set
content:
application/json:
schema:
$ref: '#/components/schemas/OkResponse'
'400':
$ref: '#/components/responses/BadRequest'
'404':
$ref: '#/components/responses/NotFound'
'500':
$ref: '#/components/responses/InternalError'
x-codeSamples:
- lang: JavaScript
source: "import Kernel from '@onkernel/sdk';\n\nconst client = new Kernel({\n apiKey: process.env['KERNEL_API_KEY'], // This is the default and can be omitted\n});\n\nconst response = await client.browsers.computer.setCursorVisibility('id', { hidden: true });\n\nconsole.log(response.ok);"
- lang: Python
source: "import os\nfrom kernel import Kernel\n\nclient = Kernel(\n api_key=os.environ.get(\"KERNEL_API_KEY\"), # This is the default and can be omitted\n)\nresponse = client.browsers.computer.set_cursor_visibility(\n id=\"id\",\n hidden=True,\n)\nprint(response.ok)"
- lang: Go
source: "package main\n\nimport (\n\t\"context\"\n\t\"fmt\"\n\n\t\"github.com/kernel/kernel-go-sdk\"\n\t\"github.com/kernel/kernel-go-sdk/option\"\n)\n\nfunc main() {\n\tclient := kernel.NewClient(\n\t\toption.WithAPIKey(\"My API Key\"),\n\t)\n\tresponse, err := client.Browsers.Computer.SetCursorVisibility(\n\t\tcontext.TODO(),\n\t\t\"id\",\n\t\tkernel.BrowserComputerSetCursorVisibilityParams{\n\t\t\tHidden: true,\n\t\t},\n\t)\n\tif err != nil {\n\t\tpanic(err.Error())\n\t}\n\tfmt.Printf(\"%+v\\n\", response.Ok)\n}\n"
/browsers/{id}/computer/batch:
post:
summary: Execute a batch of computer actions sequentially
description: 'Send an array of computer actions to execute in order on the browser instance.
Execution stops on the first error. This reduces network latency compared to
sending individual action requests.
'
operationId: batchComputerAction
tags:
- Browser Computer Controls
security:
- bearerAuth: []
parameters:
- name: id
in: path
required: true
schema:
type: string
description: Browser session ID
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/BatchComputerActionRequest'
responses:
'200':
description: All actions executed successfully
'400':
$ref: '#/components/responses/BadRequest'
'404':
$ref: '#/components/responses/NotFound'
'500':
$ref: '#/components/responses/InternalError'
x-codeSamples:
- lang: JavaScript
source: "import Kernel from '@onkernel/sdk';\n\nconst client = new Kernel({\n apiKey: process.env['KERNEL_API_KEY'], // This is the default and can be omitted\n});\n\nawait client.browsers.computer.batch('id', { actions: [{ type: 'click_mouse' }] });"
- lang: Python
source: "import os\nfrom kernel import Kernel\n\nclient = Kernel(\n api_key=os.environ.get(\"KERNEL_API_KEY\"), # This is the default and can be omitted\n)\nclient.browsers.computer.batch(\n id=\"id\",\n actions=[{\n \"type\": \"click_mouse\"\n }],\n)"
- lang: Go
source: "package main\n\nimport (\n\t\"context\"\n\n\t\"github.com/kernel/kernel-go-sdk\"\n\t\"github.com/kernel/kernel-go-sdk/option\"\n)\n\nfunc main() {\n\tclient := kernel.NewClient(\n\t\toption.WithAPIKey(\"My API Key\"),\n\t)\n\terr := client.Browsers.Computer.Batch(\n\t\tcontext.TODO(),\n\t\t\"id\",\n\t\tkernel.BrowserComputerBatchParams{\n\t\t\tActions: []kernel.BrowserComputerBatchParamsAction{{\n\t\t\t\tType: \"click_mouse\",\n\t\t\t}},\n\t\t},\n\t)\n\tif err != nil {\n\t\tpanic(err.Error())\n\t}\n}\n"
/browsers/{id}/computer/get_mouse_position:
post:
summary: Get the current mouse cursor position on the browser instance
operationId: getMousePosition
tags:
- Browser Computer Controls
security:
- bearerAuth: []
parameters:
- name: id
in: path
required: true
schema:
type: string
description: Browser session ID
responses:
'200':
description: Current mouse position
content:
application/json:
schema:
$ref: '#/components/schemas/MousePositionResponse'
'400':
$ref: '#/components/responses/BadRequest'
'404':
$ref: '#/components/responses/NotFound'
'500':
$ref: '#/components/responses/InternalError'
x-codeSamples:
- lang: JavaScript
source: "import Kernel from '@onkernel/sdk';\n\nconst client = new Kernel({\n apiKey: process.env['KERNEL_API_KEY'], // This is the default and can be omitted\n});\n\nconst response = await client.browsers.computer.getMousePosition('id');\n\nconsole.log(response.x);"
- lang: Python
source: "import os\nfrom kernel import Kernel\n\nclient = Kernel(\n api_key=os.environ.get(\"KERNEL_API_KEY\"), # This is the default and can be omitted\n)\nresponse = client.browsers.computer.get_mouse_position(\n \"id\",\n)\nprint(response.x)"
- lang: Go
source: "package main\n\nimport (\n\t\"context\"\n\t\"fmt\"\n\n\t\"github.com/kernel/kernel-go-sdk\"\n\t\"github.com/kernel/kernel-go-sdk/option\"\n)\n\nfunc main() {\n\tclient := kernel.NewClient(\n\t\toption.WithAPIKey(\"My API Key\"),\n\t)\n\tresponse, err := client.Browsers.Computer.GetMousePosition(context.TODO(), \"id\")\n\tif err != nil {\n\t\tpanic(err.Error())\n\t}\n\tfmt.Printf(\"%+v\\n\", response.X)\n}\n"
/browsers/{id}/computer/clipboard/read:
post:
summary: Read text from the clipboard on the browser instance
operationId: readClipboard
tags:
- Browser Computer Controls
security:
- bearerAuth: []
parameters:
- name: id
in: path
required: true
schema:
type: string
description: Browser session ID
responses:
'200':
description: Clipboard content read successfully
content:
application/json:
schema:
$ref: '#/components/schemas/ClipboardContent'
'400':
$ref: '#/components/responses/BadRequest'
'404':
$ref: '#/components/responses/NotFound'
'500':
$ref: '#/components/responses/InternalError'
x-codeSamples:
- lang: JavaScript
source: "import Kernel from '@onkernel/sdk';\n\nconst client = new Kernel({\n apiKey: process.env['KERNEL_API_KEY'], // This is the default and can be omitted\n});\n\nconst response = await client.browsers.computer.readClipboard('id');\n\nconsole.log(response.text);"
- lang: Python
source: "import os\nfrom kernel import Kernel\n\nclient = Kernel(\n api_key=os.environ.get(\"KERNEL_API_KEY\"), # This is the default and can be omitted\n)\nresponse = client.browsers.computer.read_clipboard(\n \"id\",\n)\nprint(response.text)"
- lang: Go
source: "package main\n\nimport (\n\t\"context\"\n\t\"fmt\"\n\n\t\"github.com/kernel/kernel-go-sdk\"\n\t\"github.com/kernel/kernel-go-sdk/option\"\n)\n\nfunc main() {\n\tclient := kernel.NewClient(\n\t\toption.WithAPIKey(\"My API Key\"),\n\t)\n\tresponse, err := client.Browsers.Computer.ReadClipboard(context.TODO(), \"id\")\n\tif err != nil {\n\t\tpanic(err.Error())\n\t}\n\tfmt.Printf(\"%+v\\n\", response.Text)\n}\n"
/browsers/{id}/computer/clipboard/write:
post:
summary: Write text to the clipboard on the browser instance
operationId: writeClipboard
tags:
- Browser Computer Controls
security:
- bearerAuth: []
parameters:
- name: id
in: path
required: true
schema:
type: string
description: Browser session ID
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/WriteClipboardRequest'
responses:
'200':
description: Text written to clipboard successfully
'400':
$ref: '#/components/responses/BadRequest'
'404':
$ref: '#/components/responses/NotFound'
'500':
$ref: '#/components/responses/InternalError'
x-codeSamples:
- lang: JavaScript
source: "import Kernel from '@onkernel/sdk';\n\nconst client = new Kernel({\n apiKey: process.env['KERNEL_API_KEY'], // This is the default and can be omitted\n});\n\nawait client.browsers.computer.writeClipboard('id', { text: 'text' });"
- lang: Python
source: "import os\nfrom kernel import Kernel\n\nclient = Kernel(\n api_key=os.environ.get(\"KERNEL_API_KEY\"), # This is the default and can be omitted\n)\nclient.browsers.computer.write_clipboard(\n id=\"id\",\n text=\"text\",\n)"
- lang: Go
source: "package main\n\nimport (\n\t\"context\"\n\n\t\"github.com/kernel/kernel-go-sdk\"\n\t\"github.com/kernel/kernel-go-sdk/option\"\n)\n\nfunc main() {\n\tclient := kernel.NewClient(\n\t\toption.WithAPIKey(\"My API Key\"),\n\t)\n\terr := client.Browsers.Computer.WriteClipboard(\n\t\tcontext.TODO(),\n\t\t\"id\",\n\t\tkernel.BrowserComputerWriteClipboardParams{\n\t\t\tText: \"text\",\n\t\t},\n\t)\n\tif err != nil {\n\t\tpanic(err.Error())\n\t}\n}\n"
components:
schemas:
ScreenshotRequest:
type: object
properties:
region:
$ref: '#/components/schemas/ScreenshotRegion'
additionalProperties: false
SetCursorRequest:
type: object
required:
- hidden
properties:
hidden:
type: boolean
description: Whether the cursor should be hidden or visible
additionalProperties: false
PressKeyRequest:
type: object
required:
- keys
properties:
keys:
type: array
description: 'List of key symbols to press. Each item should be a key symbol supported by xdotool
(see X11 keysym definitions). Examples include "Return", "Shift", "Ctrl", "Alt", "F5".
Items in this list could also be combinations, e.g. "Ctrl+t" or "Ctrl+Shift+Tab".
'
items:
type: string
duration:
type: integer
description: Duration to hold the keys down in milliseconds. If omitted or 0, keys are tapped.
minimum: 0
default: 0
hold_keys:
type: array
description: Optional modifier keys to hold during the key press sequence.
items:
type: string
additionalProperties: false
ScreenshotRegion:
type: object
required:
- x
- y
- width
- height
properties:
x:
type: integer
description: X coordinate of the region's top-left corner
y:
type: integer
description: Y coordinate of the region's top-left corner
width:
type: integer
description: Width of the region in pixels
height:
type: integer
description: Height of the region in pixels
additionalProperties: false
ClipboardContent:
type: object
required:
- text
properties:
text:
type: string
description: Current clipboard text content
additionalProperties: false
MousePositionResponse:
type: object
required:
- x
- y
properties:
x:
type: integer
description: X coordinate of the cursor
y:
type: integer
description: Y coordinate of the cursor
additionalProperties: false
SleepAction:
type: object
description: Pause execution for a specified duration.
required:
- duration_ms
properties:
duration_ms:
type: integer
description: Duration to sleep in milliseconds.
minimum: 0
maximum: 30000
additionalProperties: false
WriteClipboardRequest:
type: object
required:
- text
properties:
text:
type: string
description: Text to write to the system clipboard
additionalProperties: false
ErrorDetail:
type: object
properties:
code:
type: string
description: Lower-level error code providing more specific detail
example: invalid_input
message:
type: string
description: Further detail about the error
example: Provided version string is not semver compliant
ScrollRequest:
type: object
required:
- x
- y
properties:
x:
type: integer
description: X coordinate at which to perform the scroll
y:
type: integer
description: Y coordinate at which to perform the scroll
delta_x:
type: integer
description: Horizontal scroll amount in xdotool "wheel units." Positive scrolls right, negative scrolls left.
default: 0
delta_y:
type: integer
description: Vertical scroll amount in xdotool "wheel units." Positive scrolls down, negative scrolls up.
default: 0
hold_keys:
type: array
description: Modifier keys to hold during the scroll
items:
type: string
additionalProperties: false
Error:
type: object
required:
- code
- message
properties:
code:
type: string
description: Application-specific error code (machine-readable)
example: bad_request
message:
type: string
description: Human-readable error description for debugging
example: 'Missing required field: app_name'
details:
type: array
description: Additional error details (for multiple errors)
items:
$ref: '#/components/schemas/ErrorDetail'
inner_error:
$ref: '#/components/schemas/ErrorDetail'
TypeTextRequest:
type: object
required:
- text
properties:
text:
type: string
description: Text to type on the browser instance
delay:
type: integer
description: Delay in milliseconds between keystrokes
minimum: 0
default: 0
additionalProperties: false
ComputerAction:
type: object
description: 'A single computer action to execute as part of a batch. The `type` field selects which
action to perform, and the corresponding field contains the action parameters.
Exactly one action field matching the type must be provided.
'
required:
- type
properties:
type:
type: string
description: The type of action to perform.
enum:
- click_mouse
- move_mouse
- type_text
- press_key
- scroll
- drag_mouse
- set_cursor
- sleep
click_mouse:
$ref: '#/components/schemas/ClickMouseRequest'
move_mouse:
$ref: '#/components/schemas/MoveMouseRequest'
type_text:
$ref: '#/components/schemas/TypeTextRequest'
press_key:
$ref: '#/components/schemas/PressKeyRequest'
scroll:
$ref: '#/components/schemas/ScrollRequest'
drag_mouse:
$ref: '#/components/schemas/DragMouseRequest'
set_cursor:
$ref: '#/components/schemas/SetCursorRequest'
sleep:
$ref: '#/components/schemas/SleepAction'
additionalProperties: false
ClickMouseRequest:
type: object
required:
- x
- y
properties:
button:
type: string
description: Mouse button to interact with
enum:
- left
- right
- middle
- back
- forward
click_type:
type: string
description: Type of click action
enum:
- down
- up
- click
x:
type: integer
description: X coordinate of the click position
y:
type: integer
description: Y coordinate of the click position
hold_keys:
type: array
description: Modifier keys to hold during the click
items:
type: string
num_clicks:
type: integer
description: Number of times to repeat the click
default: 1
additionalProperties: false
DragMouseRequest:
type: object
required:
- path
properties:
path:
type: array
description: Ordered list of [x, y] coordinate pairs to move through while dragging. Must contain at least 2 points.
minItems: 2
items:
type: array
minItems: 2
maxItems: 2
items:
type: integer
button:
type: string
description: Mouse button to drag with
enum:
- left
- middle
- right
delay:
type: integer
description: Delay in milliseconds between button down and starting to move along the path.
minimum: 0
default: 0
steps_per_segment:
type: integer
description: Number of relative move steps per segment in the path. Minimum 1.
minimum: 1
default: 10
step_delay_ms:
type: integer
description: Delay in milliseconds between relative steps while dragging (not the initial delay).
minimum: 0
default: 50
hold_keys:
type: array
description: Modifier keys to hold during the drag
items:
type: string
smooth:
type: boolean
description: Use human-like Bezier curves between path waypoints instead of linear interpolation. When true, steps_per_segment and step_delay_ms are ignored.
default: true
duration_ms:
type: integer
description: Target total duration in milliseconds for the entire drag movement when smooth=true. Omit for automatic
# --- truncated at 32 KB (34 KB total) ---
# Full source: https://raw.githubusercontent.com/api-evangelist/kernel/refs/heads/main/openapi/kernel-browser-computer-controls-api-openapi.yml