ReactorSDK

reactor_sdk is a Ruby SDK for the Adobe Launch / Data Collection Reactor API. It handles OAuth Server-to-Server authentication, JSON:API parsing, pagination, retries, rate limiting, revision snapshots, and review-friendly comparison helpers so application code can work with typed Ruby objects instead of raw HTTP payloads.

Features

OAuth Server-to-Server authentication with configurable automatic token refresh
Typed resource objects across companies, properties, rules, data elements, extensions, libraries, builds, revisions, notes, and more
Automatic pagination for list endpoints
Consistent error classes and retry behavior around Adobe API failures
Library-aware review helpers for rules, data elements, and extensions
Upstream-chain lookup for resource review across Development, Staging, and Production
Library-to-library comparison output with status, revision IDs, and normalized review payloads

Installation

Add the gem to your application:

gem "reactor_sdk"

Then install it:

bundle install

Or install directly:

gem install reactor_sdk

Authentication and Configuration

ReactorSDK uses Adobe OAuth Server-to-Server credentials. JWT / Service Account credentials are not supported.

require "reactor_sdk"

client = ReactorSDK::Client.new(
  client_id: ENV.fetch("ADOBE_CLIENT_ID"),
  client_secret: ENV.fetch("ADOBE_CLIENT_SECRET"),
  org_id: ENV.fetch("ADOBE_IMS_ORG_ID")
)

Optional configuration is passed to the same client constructor:

client = ReactorSDK::Client.new(
  client_id: "your-client-id",
  client_secret: "your-client-secret",
  org_id: "your-org-id@AdobeOrg",
  base_url: "https://reactor.adobe.io",
  ims_token_url: "https://ims-na1.adobelogin.com/ims/token/v3",
  timeout: 30,
  logger: Logger.new($stdout),
  auto_refresh_token: true
)

auto_refresh_token defaults to true. When you set it to false, ReactorSDK still fetches the initial Adobe IMS token, but once that cached token expires the SDK raises ReactorSDK::AuthenticationError instead of silently refreshing it.

Quick Start

require "reactor_sdk"

client = ReactorSDK::Client.new(
  client_id: ENV.fetch("ADOBE_CLIENT_ID"),
  client_secret: ENV.fetch("ADOBE_CLIENT_SECRET"),
  org_id: ENV.fetch("ADOBE_IMS_ORG_ID")
)

# List companies available to the credential set.
companies = client.companies.list
company = companies.first

# List properties inside the company.
properties = client.properties.list_for_company(company.id)
property = properties.first

# List rules in the property.
rules = client.rules.list_for_property(property.id)
puts rules.first&.name

Review and Promotion Workflows

ReactorSDK includes higher-level helpers for review tooling and release workflows, not just CRUD wrappers.

Work with a Library Snapshot

# LB_DEV = development library ID.
# PR123 = property ID.
snapshot = client.libraries.find_snapshot("LB_DEV", property_id: "PR123")

# Point-in-time rule components associated with the rule inside this library snapshot.
snapshot.rule_components_for_rule("RL123").map(&:id)

# Snapshot-scoped impact analysis for a data element.
snapshot.impacted_rules_for("DE123").map(&:name)

find_snapshot returns the effective Launch snapshot for that library, including inherited upstream resources that are still active for review and build decisions. If you need only the records directly attached to one library, call find_direct_snapshot instead:

direct_snapshot = client.libraries.find_direct_snapshot(
  "LB_DEV",
  property_id: "PR123"
)

Build Comprehensive Review Objects

rule_review = client.rules.find_comprehensive(
  "RL123",
  library_id: "LB_DEV",
  property_id: "PR123"
)

data_element_review = client.data_elements.find_comprehensive(
  "DE123",
  library_id: "LB_DEV",
  property_id: "PR123"
)

extension_review = client.extensions.find_comprehensive(
  "EX123",
  library_id: "LB_DEV",
  property_id: "PR123"
)

puts rule_review.normalized_json
puts data_element_review.normalized_json
puts extension_review.normalized_json

These comprehensive review objects include the resource plus the associated records a reviewer usually needs:

Rules include rule components
Data elements include referenced data elements and impacted rules
Extensions include dependent data elements, rule components, and rules

Resolve Upstream Versions

chain = client.rules.comprehensive_upstream_chain(
  "RL123",
  library_id: "LB_DEV",
  property_id: "PR123"
)

staging_entry = chain.entries.find { |entry| entry.stage == "staging" }

puts chain.target_revision_id
puts staging_entry&.revision_id
puts staging_entry&.normalized_json

Compare Two Libraries

# Compare Development against Staging.
comparison = client.libraries.compare(
  "LB_DEV",
  baseline_library_id: "LB_STG",
  property_id: "PR123"
)

comparison.entries.each do |entry|
  puts "#{entry.resource_type} #{entry.resource_id} #{entry.status}"
  puts entry.current_revision_id
  puts entry.baseline_revision_id
end

# Inspect normalized output for one entry.
entry = comparison.entries.first
puts entry.current_normalized_json
puts entry.baseline_normalized_json

Sample comparison output:

# Example status rows you might print from comparison.entries:
# rules RL100 modified
# current revision  = RE_CUR_RULE
# baseline revision = RE_BASE_RULE
#
# rules RL200 added
# current revision  = RE_ADDED
# baseline revision = nil
#
# data_elements DE200 unchanged
# current revision  = RE_REMOVED
# baseline revision = RE_REMOVED
#
# extensions EX100 unchanged
# current revision  = RE_EX
# baseline revision = RE_EX

Comparison is performed across effective library snapshots, so unchanged resources inherited from Staging or Production are preserved as unchanged rather than being misreported as removed. Rule components are still included in the comprehensive rule payload so rule diffs remain readable.

Migration Notes for 1.0.0

find_snapshot now returns the effective inherited Launch state for a library. If your application relied on the old direct-only behavior, switch that code to find_direct_snapshot.
compare now evaluates effective inherited snapshots, so unchanged upstream resources may move from removed to unchanged in review output.
auto_refresh_token: false is now enforced. If your app sets it to false, be ready to recreate clients or handle ReactorSDK::AuthenticationError after token expiry.

Endpoint Coverage

The SDK currently exposes these endpoint families:

Companies
Properties
App configurations
Callbacks
Secrets
Environments
Hosts
Rules
Rule components
Data elements
Extensions
Extension packages
Extension package usage authorizations
Libraries
Builds
Revisions
Notes
Audit events
Profiles
Search

Documentation

Full guides, API reference pages, example payloads, and review-workflow documentation are available at:

https://reactor-sdk.dhairyagabhawala.com

Development

Install dependencies:

bundle install

Run the test suite:

bundle exec rspec

Run RuboCop:

bundle exec rubocop

Build the gem:

gem build reactor_sdk.gemspec

Contributing

See CONTRIBUTING.md for contribution and development guidance.