roast-ai 1.2.0 → 1.4.0

data/tutorial/01_your_first_workflow/README.md

@@ -106,7 +106,7 @@ block:
 config do
   chat do
     model "gpt-4o-mini" # Use OpenAI's fast model
-    provider :openai # Use OpenAI (can also be :anthropic, :perplexity or :gemini)
+    provider :openai # Use OpenAI (can also be :anthropic, :perplexity, :gemini or :bedrock)
     show_prompt! # Display the prompt before sending
   end
 end
@@ -129,8 +129,9 @@ Common options you can set:
     - Anthropic: "claude-haiku-4-5" (default), "claude-sonnet-4-6", "claude-opus-4-7", etc.
     - Perplexity: "sonar" (default), "sonar-pro", "sonar-deep-research", etc.
     - Gemini: "gemini-3.1-flash-lite" (default), "gemini-3-flash-preview", "gemini-3.1-pro-preview", etc.
+    - Bedrock: "us.anthropic.claude-haiku-4-5-20251001-v1:0" (default), or any other Bedrock model or inference profile ID your account can use
 - `provider :name` - Which LLM provider
-    - `:openai`, `:anthropic`, `:perplexity` or `:gemini`
+    - `:openai`, `:anthropic`, `:perplexity`, `:gemini` or `:bedrock`
 - `show_prompt!` - Display the prompt being sent
 - `show_response!` - Display the response (on by default)
 - `show_stats!` - Display token usage statistics (on by default)

data/tutorial/01_your_first_workflow/configured_chat.rb

@@ -10,7 +10,7 @@ config do
   # Configure all chat cogs in this workflow
   chat do
     model "gpt-4o-mini"      # Use OpenAI's fast, cost-effective model
-    provider :openai         # Use OpenAI (alternative: :anthropic, :perplexity or :gemini)
+    provider :openai         # Use OpenAI (alternative: :anthropic, :perplexity, :gemini or :bedrock)
     show_prompt!             # Display the prompt before sending it
     show_response!           # Display the response (this is the default)
     show_stats!              # Display token usage statistics (default)

data/tutorial/02_chaining_cogs/README.md

@@ -127,8 +127,29 @@ execute do
 end
 ```
 
-The `agent` cog is backed by a locally installed coding agent -- Anthropic's Claude Code is the default provider.
-You'll need to have Claude Code installed and configured correctly for this cog to run.
+The `agent` cog is backed by a locally installed coding agent -- Pi is the default provider (set `ROAST_DEFAULT_AGENT_PROVIDER` to change it globally, e.g. to `claude`).
+You'll need to have Pi installed and configured correctly for this cog to run.
+Roast also supports Claude code, which must be set explicitly in provider or an environment variable.
+
+### Specifying a Model
+
+Use a fully-qualified `provider/model` id:
+
+```ruby
+config do
+  agent do
+    model "anthropic/claude-haiku-4-5-20251001"
+  end
+end
+```
+
+Pi requires this qualified form. Claude Code expects a bare model name, so Roast strips the leading
+`anthropic/` prefix automatically when the `:claude` provider is used. Because of this, the same `model`
+string is portable across providers -- you can switch the agent provider without rewriting it.
+
+A bare name like `"claude-haiku-4-5-20251001"` (or a shorthand like `"haiku"`) still works when you've set
+`provider :claude`, but it will fail on the default Pi provider, which can't resolve an unqualified model name.
+Prefer the qualified form unless you have a reason not to.
 
 ### When to Use Agent vs Chat
 

data/tutorial/02_chaining_cogs/code_review.rb

@@ -8,8 +8,8 @@
 
 config do
   agent do
-    model "claude-haiku-4-5-20251001"
-    provider :claude
+    model "anthropic/claude-haiku-4-5-20251001"
+    provider :pi
   end
 
   chat do

data/tutorial/02_chaining_cogs/session_resumption.rb

@@ -13,6 +13,7 @@ config do
     model "gpt-5.4-nano"
   end
   agent do
+    provider :claude
     model "haiku"
     no_display!
     show_stats!

data/tutorial/03_targets_and_params/README.md

@@ -187,6 +187,15 @@ execute do
 end
 ```
 
+Parameter accessors are also available inside cog `config` blocks and the `global` block:
+
+```ruby
+config do
+  chat { model(arg?(:fast) ? "gpt-5.4-nano" : "gpt-5") }
+  global { abort_on_failure! if arg?(:strict) }
+end
+```
+
 ## Running the Workflows
 
 To run the examples in this chapter:
@@ -220,7 +229,7 @@ bin/roast execute tutorial/03_targets_and_params/multiple_targets.rb \
 - Use `arg?(:name)` to check if an argument is present
 - Use `kwarg(:name)` to get a kwarg value (returns nil if missing)
 - Use `kwarg!(:name)` to require a kwarg (errors if missing)
-- All accessors work in any cog's input block
+- All accessors work in any cog's input block, and inside cog `config` / `global` blocks
 
 ## What's Next?
 

data/tutorial/04_configuration_options/README.md

@@ -111,6 +111,25 @@ end
 
 This is useful when you have multiple similar cogs that need the same configuration.
 
+### Configuring Based on Parameters
+
+Workflow parameter accessors (`target!`, `targets`, `arg?`, `args`, `kwarg`, `kwarg!`, `kwarg?`, `kwargs`; see Chapter 3)
+are available inside cog `config` blocks and the `global` block, so you can configure cogs based on how the workflow was
+invoked:
+
+```ruby
+config do
+  chat do
+    # Pick a cheaper model when invoked with the `fast` flag: `-- fast`
+    model(arg?(:fast) ? "gpt-5.4-nano" : "gpt-5")
+  end
+
+  global do
+    abort_on_failure! if arg?(:strict)
+  end
+end
+```
+
 ## Display Options
 
 Control what gets printed during workflow execution using display methods in your `config` block:

data/tutorial/README.md

@@ -15,7 +15,7 @@ run coding agents, process data, and so much more.
 - Ruby installed (3.4.2+)
 - Roast gem installed
 - API keys for your AI provider (see [Configuration](https://github.com/Shopify/roast/blob/main/README.md#configuration) for setup)
-- Claude Code CLI installed and configured (to use coding agents)
+- Pi CLI installed and configured for default agent examples; Claude Code CLI installed and configured for examples that set `provider :claude`
 
 ## How to Use This Tutorial
 

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: roast-ai
 version: !ruby/object:Gem::Version
-  version: 1.2.0
+  version: 1.4.0
 platform: ruby
 authors:
 - Shopify
@@ -245,6 +245,7 @@ files:
 - lib/roast/version.rb
 - lib/roast/workflow.rb
 - lib/roast/workflow_context.rb
+- lib/roast/workflow_param_accessors.rb
 - lib/roast/workflow_params.rb
 - roast-ai.gemspec
 - shipit.rubygems.yml
@@ -349,6 +350,7 @@ files:
 - sorbet/rbi/gems/vcr@6.3.1.rbi
 - sorbet/rbi/gems/webmock@3.25.1.rbi
 - sorbet/rbi/gems/zeitwerk@2.7.3.rbi
+- sorbet/rbi/shims/lib/roast/cog/config.rbi
 - sorbet/rbi/shims/lib/roast/cog_input_context.rbi
 - sorbet/rbi/shims/lib/roast/config_context.rbi
 - sorbet/rbi/shims/lib/roast/execution_context.rbi
@@ -406,7 +408,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 4.0.14
+rubygems_version: 4.0.22
 specification_version: 4
 summary: A framework for executing structured AI workflows in Ruby
 test_files: []