Skip to main content
This documentation is automatically synced from the Ruby SDK GitHub repository.
The Stagehand Ruby library provides convenient access to the Stagehand REST API from any Ruby 3.2.0+ application. It ships with comprehensive types & docstrings in Yard, RBS, and RBI – see below for usage with Sorbet. The standard library’s net/http is used as the HTTP transport, with connection pooling via the connection_pool gem. It is generated with Stainless.

​
Documentation

Documentation for releases of this gem can be found on RubyDoc. The REST API documentation can be found on docs.stagehand.dev.

​
Installation

To use this gem, install via Bundler by adding the following to your application’s Gemfile: 
gem "stagehand", "~> 0.6.0"

​
Usage

require "bundler/setup"
require "stagehand"

# Create a new Stagehand client with your credentials
client = Stagehand::Client.new(
  browserbase_api_key: ENV["BROWSERBASE_API_KEY"],      # defaults to ENV["BROWSERBASE_API_KEY"]
  browserbase_project_id: ENV["BROWSERBASE_PROJECT_ID"], # defaults to ENV["BROWSERBASE_PROJECT_ID"]
  model_api_key: ENV["MODEL_API_KEY"]                   # defaults to ENV["MODEL_API_KEY"]
)

# Start a new browser session
# x_language and x_sdk_version headers are required for the v3 API
start_response = client.sessions.start(
  model_name: "openai/gpt-5-nano",
  x_language: :typescript,
  x_sdk_version: "3.0.6"
)
puts "Session started: #{start_response.data.session_id}"

session_id = start_response.data.session_id

# Navigate to a webpage
# frame_id is required - use empty string for the main frame
client.sessions.navigate(
  session_id,
  url: "https://news.ycombinator.com",
  frame_id: "",
  x_language: :typescript,
  x_sdk_version: "3.0.6"
)
puts "Navigated to Hacker News"

# Use Observe to find possible actions on the page
observe_response = client.sessions.observe(
  session_id,
  instruction: "find the link to view comments for the top post",
  x_language: :typescript,
  x_sdk_version: "3.0.6"
)

actions = observe_response.data.result
puts "Found #{actions.length} possible actions"

# Take the first action returned by Observe
action = actions.first
puts "Acting on: #{action.description}"

# Pass the structured action to Act
# Convert the observe result to a hash and ensure method is set to "click"
act_response = client.sessions.act(
  session_id,
  input: action.to_h.merge(method: "click"),
  x_language: :typescript,
  x_sdk_version: "3.0.6"
)
puts "Act completed: #{act_response.data.result[:message]}"

# Extract data from the page
# We're now on the comments page, so extract the top comment text
extract_response = client.sessions.extract(
  session_id,
  instruction: "extract the text of the top comment on this page",
  schema: {
    type: "object",
    properties: {
      comment_text: {
        type: "string",
        description: "The text content of the top comment"
      },
      author: {
        type: "string",
        description: "The username of the comment author"
      }
    },
    required: ["comment_text"]
  },
  x_language: :typescript,
  x_sdk_version: "3.0.6"
)
puts "Extracted data: #{extract_response.data.result}"

# Get the author from the extracted data
extracted_data = extract_response.data.result
author = extracted_data[:author]
puts "Looking up profile for author: #{author}"

# Use the Agent to find the author's profile
# Execute runs an autonomous agent that can navigate and interact with pages
execute_response = client.sessions.execute(
  session_id,
  execute_options: {
    instruction: "Find any personal website, GitHub, LinkedIn, or other best profile URL for the Hacker News user '#{author}'. " \
                 "Click on their username to go to their profile page and look for any links they have shared.",
    max_steps: 15
  },
  agent_config: {
    model: Stagehand::ModelConfig::ModelConfigObject.new(
      model_name: "openai/gpt-5-nano",
      api_key: ENV["MODEL_API_KEY"]
    ),
    cua: false
  },
  x_language: :typescript,
  x_sdk_version: "3.0.6"
)
puts "Agent completed: #{execute_response.data.result[:message]}"
puts "Agent success: #{execute_response.data.result[:success]}"
puts "Agent actions taken: #{execute_response.data.result[:actions]&.length || 0}"

# End the session to cleanup browser resources
client.sessions.end_(
  session_id,
  x_language: :typescript,
  x_sdk_version: "3.0.6"
)
puts "Session ended"

​
Running the Example

Set the required environment variables and run the example script: 
# Set your credentials
export BROWSERBASE_API_KEY="your-browserbase-api-key"
export BROWSERBASE_PROJECT_ID="your-browserbase-project-id"
export MODEL_API_KEY="your-openai-api-key"

# Install dependencies and run
bundle install
bundle exec ruby examples/basic.rb

​
Streaming

We provide support for streaming responses using Server-Sent Events (SSE). 
stream = stagehand.sessions.act_streaming(
  "00000000-your-session-id-000000000000",
  input: "click the first link on the page"
)

stream.each do |session|
  puts(session.data)
end

​
Handling errors

When the library is unable to connect to the API, or if the API returns a non-success status code (i.e., 4xx or 5xx response), a subclass of Stagehand::Errors::APIError will be thrown: 
begin
  session = stagehand.sessions.start(model_name: "openai/gpt-5-nano")
rescue Stagehand::Errors::APIConnectionError => e
  puts("The server could not be reached")
  puts(e.cause)  # an underlying Exception, likely raised within `net/http`
rescue Stagehand::Errors::RateLimitError => e
  puts("A 429 status code was received; we should back off a bit.")
rescue Stagehand::Errors::APIStatusError => e
  puts("Another non-200-range status code was received")
  puts(e.status)
end
Error codes are as follows:
CauseError Type
HTTP 400BadRequestError
HTTP 401AuthenticationError
HTTP 403PermissionDeniedError
HTTP 404NotFoundError
HTTP 409ConflictError
HTTP 422UnprocessableEntityError
HTTP 429RateLimitError
HTTP >= 500InternalServerError
Other HTTP errorAPIStatusError
TimeoutAPITimeoutError
Network errorAPIConnectionError

​
Retries

Certain errors will be automatically retried 2 times by default, with a short exponential backoff. Connection errors (for example, due to a network connectivity problem), 408 Request Timeout, 409 Conflict, 429 Rate Limit, >=500 Internal errors, and timeouts will all be retried by default. You can use the max_retries option to configure or disable this: 
# Configure the default for all requests:
stagehand = Stagehand::Client.new(
  max_retries: 0 # default is 2
)

# Or, configure per-request:
stagehand.sessions.start(model_name: "openai/gpt-5-nano", request_options: {max_retries: 5})

​
Timeouts

By default, requests will time out after 60 seconds. You can use the timeout option to configure or disable this: 
# Configure the default for all requests:
stagehand = Stagehand::Client.new(
  timeout: nil # default is 60
)

# Or, configure per-request:
stagehand.sessions.start(model_name: "openai/gpt-5-nano", request_options: {timeout: 5})
On timeout, Stagehand::Errors::APITimeoutError is raised. Note that requests that time out are retried by default.

​
Advanced concepts

​
BaseModel

All parameter and response objects inherit from Stagehand::Internal::Type::BaseModel, which provides several conveniences, including:
  1. All fields, including unknown ones, are accessible with obj[:prop] syntax, and can be destructured with obj => {prop: prop} or pattern-matching syntax.
  2. Structural equivalence for equality; if two API calls return the same values, comparing the responses with == will return true.
  3. Both instances and the classes themselves can be pretty-printed.
  4. Helpers such as #to_h, #deep_to_h, #to_json, and #to_yaml.

​
Making custom or undocumented requests

​
Undocumented properties

You can send undocumented parameters to any endpoint, and read undocumented response properties, like so: Note: the extra_ parameters of the same name overrides the documented parameters. 
response =
  stagehand.sessions.start(
    model_name: "openai/gpt-5-nano",
    request_options: {
      extra_query: {my_query_parameter: value},
      extra_body: {my_body_parameter: value},
      extra_headers: {"my-header": value}
    }
  )

puts(response[:my_undocumented_property])

​
Undocumented request params

If you want to explicitly send an extra param, you can do so with the extra_query, extra_body, and extra_headers under the request_options: parameter when making a request, as seen in the examples above.

​
Undocumented endpoints

To make requests to undocumented endpoints while retaining the benefit of auth, retries, and so on, you can make requests using client.request, like so: 
response = client.request(
  method: :post,
  path: '/undocumented/endpoint',
  query: {"dog": "woof"},
  headers: {"useful-header": "interesting-value"},
  body: {"hello": "world"}
)

​
Concurrency & connection pooling

The Stagehand::Client instances are threadsafe, but are only are fork-safe when there are no in-flight HTTP requests. Each instance of Stagehand::Client has its own HTTP connection pool with a default size of 99. As such, we recommend instantiating the client once per application in most settings. When all available connections from the pool are checked out, requests wait for a new connection to become available, with queue time counting towards the request timeout. Unless otherwise specified, other classes in the SDK do not have locks protecting their underlying data structure.

​
Sorbet

This library provides comprehensive RBI definitions, and has no dependency on sorbet-runtime. You can provide typesafe request parameters like so: 
stagehand.sessions.act("00000000-your-session-id-000000000000", input: "click the first link on the page")
Or, equivalently: 
# Hashes work, but are not typesafe:
stagehand.sessions.act("00000000-your-session-id-000000000000", input: "click the first link on the page")

# You can also splat a full Params class:
params = Stagehand::SessionActParams.new(input: "click the first link on the page")
stagehand.sessions.act("00000000-your-session-id-000000000000", **params)

​
Enums

Since this library does not depend on sorbet-runtime, it cannot provide T::Enum instances. Instead, we provide “tagged symbols” instead, which is always a primitive at runtime: 
# :typescript
puts(Stagehand::SessionActParams::XLanguage::TYPESCRIPT)

# Revealed type: `T.all(Stagehand::SessionActParams::XLanguage, Symbol)`
T.reveal_type(Stagehand::SessionActParams::XLanguage::TYPESCRIPT)
Enum parameters have a “relaxed” type, so you can either pass in enum constants or their literal value: 
# Using the enum constants preserves the tagged type information:
stagehand.sessions.act(
  x_language: Stagehand::SessionActParams::XLanguage::TYPESCRIPT,
  # …
)

# Literal values are also permissible:
stagehand.sessions.act(
  x_language: :typescript,
  # …
)

​
Versioning

This package follows SemVer conventions. As the library is in initial development and has a major version of 0, APIs may change at any time. This package considers improvements to the (non-runtime) *.rbi and *.rbs type definitions to be non-breaking changes.

​
Requirements

Ruby 3.2.0 or higher.

​
Contributing

See the contributing documentation.