lex-llm-vertex 0.2.8 → 0.2.10

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: d7dbca12cb0f99264ee9fd713291622b0ee9ad6f238193811bdcfb2847be3211
-  data.tar.gz: be61eafce74bdaec30c737ec906abafbbacad3f4c125b52199db8de8dd3d3ca6
+  metadata.gz: 25083d8e7cc928d57127feab13406395a6600421913559042fecfe7e2a376267
+  data.tar.gz: '028cb717f3299a2683bdd0ed10cc59f91a75f3de5f773d730ec75e2374762170'
 SHA512:
-  metadata.gz: f557d2fbf831656f3aa0d4a0788ecfedab7889f5d9fdd666300ddad2be3348184e56c8bc0f5d3a469b64d9d12e2fae846bcfbd97100ff6a8b1f755cdaea803c2
-  data.tar.gz: 90199aa13e8d105931040d6131f9c1591cc7bc956dd61e0046589a7eba06d0ab5aae04cd8fdce276a28adafd980e8c3a4a781687bc626e31097a719a8c47220b
+  metadata.gz: 69f44a1e8ac1d5d0da6b9e4ca5b5d5f7f0fba37b49021e5a512db86884958a89abd20323c336b9292c1973fb47e303d37b715322aaf37da75ab1f3d5da33a48f
+  data.tar.gz: 53cb2cfeb3039b78ca84722943e220178b6a7c1a060972ca329aacd4acb742bac4e4a581d9df819fa1902f315ca73858d9401d29ec08318c7ce0290899b11fc7

data/CHANGELOG.md

@@ -1,5 +1,18 @@
 # Changelog
 
+## 0.2.10 - 2026-06-02
+
+- Add per-provider scoped discovery refresh actor
+
+## 0.2.9 - 2026-05-21
+
+- Add `default_transport`/`default_tier` class declarations, remove `configured_transport`/`configured_tier`
+- Remove `DEFAULT_LOCATION`/`DEFAULT_PROJECT`/`DEFAULT_PUBLISHER` constants — now read from settings
+- Add `model_allowed?` filtering in `discover_offerings`
+- Default tier corrected from :frontier to :cloud
+- Identity headers included via base provider
+
+
 ## 0.2.8 - 2026-05-18
 
 - Fix streaming tool calls: `build_chunk` now passes `tool_calls: parse_tool_calls(parts)` to the Chunk constructor. Previously tool calls were omitted from streaming responses entirely.

data/lib/legion/extensions/llm/vertex/actors/discovery_refresh.rb

@@ -0,0 +1,47 @@
+# frozen_string_literal: true
+
+begin
+  require 'legion/extensions/actors/every'
+rescue LoadError => e
+  warn(e.message) if $VERBOSE
+end
+
+return unless defined?(Legion::Extensions::Actors::Every)
+
+module Legion
+  module Extensions
+    module Llm
+      module Vertex
+        module Actor
+          class DiscoveryRefresh < Legion::Extensions::Actors::Every # rubocop:disable Style/Documentation
+            include Legion::Logging::Helper
+
+            REFRESH_INTERVAL = 1800
+
+            def runner_class    = self.class
+            def runner_function = 'manual'
+            def run_now?        = true
+            def use_runner?     = false
+            def check_subtask?  = false
+            def generate_task?  = false
+
+            def time
+              return REFRESH_INTERVAL unless defined?(Legion::Settings)
+
+              Legion::Settings.dig(:extensions, :llm, :vertex, :discovery_interval) || REFRESH_INTERVAL
+            end
+
+            def manual
+              log.debug('[vertex][discovery_refresh] refreshing model list')
+              return unless defined?(Legion::LLM::Discovery)
+
+              Legion::LLM::Discovery.refresh_discovered_models!(provider: :vertex)
+            rescue StandardError => e
+              handle_exception(e, level: :warn, handled: true, operation: 'vertex.actor.discovery_refresh')
+            end
+          end
+        end
+      end
+    end
+  end
+end

data/lib/legion/extensions/llm/vertex/provider.rb

@@ -11,10 +11,6 @@ module Legion
       module Vertex
         # Google Cloud Vertex AI provider implementation for the Legion::Extensions::Llm contract.
         class Provider < Legion::Extensions::Llm::Provider # rubocop:disable Metrics/ClassLength
-          DEFAULT_LOCATION = 'us-central1'
-          DEFAULT_PROJECT = 'env://GOOGLE_CLOUD_PROJECT'
-          DEFAULT_PUBLISHER = 'google'
-
           STATIC_MODELS = [
             { model: 'gemini-2.5-flash', alias: 'gemini-flash', publisher: 'google', model_family: :gemini },
             { model: 'gemini-2.5-pro', alias: 'gemini-pro', publisher: 'google', model_family: :gemini },
@@ -39,6 +35,8 @@ module Legion
             attr_writer :registry_publisher
 
             def slug = 'vertex'
+            def default_transport = :http
+            def default_tier = :cloud
 
             def configuration_options
               %i[
@@ -83,16 +81,22 @@ module Legion
             end
           end
 
+          def settings
+            Vertex.default_settings
+          end
+
           def api_base
             config.vertex_api_base || "https://#{location}-aiplatform.googleapis.com/v1"
           end
 
           def headers
-            { 'Authorization' => bearer_token, 'Content-Type' => 'application/json; charset=utf-8' }.compact
+            identity_headers.merge({ 'Authorization' => bearer_token,
+                                     'Content-Type' => 'application/json; charset=utf-8' }.compact)
           end
 
-          def project = config.vertex_project || ENV.fetch('GOOGLE_CLOUD_PROJECT', DEFAULT_PROJECT)
-          def location = config.vertex_location || DEFAULT_LOCATION
+          def project = config.vertex_project || settings[:project] || ENV.fetch('GOOGLE_CLOUD_PROJECT', nil)
+          def location = config.vertex_location || settings[:location] || 'us-central1'
+          def default_publisher = settings[:publisher] || 'google'
           def models_url = publisher_parent
           def completion_url = generate_content_url(model: @model || STATIC_MODELS.first.fetch(:model))
           def stream_url = stream_generate_content_url(model: @model || STATIC_MODELS.first.fetch(:model))
@@ -126,7 +130,13 @@ module Legion
 
             response = connection.get(models_url)
             models = response.body['publisherModels'] || response.body['models'] || []
-            offerings = models.map { |model| offering_from_live_model(model) }
+            offerings = models.filter_map do |model|
+              offering = offering_from_live_model(model)
+              model_id = offering.respond_to?(:model) ? offering.model : (offering[:model] || offering[:id])
+              next unless model_allowed?(model_id.to_s)
+
+              offering
+            end
             log.info { "discovered #{offerings.size} live offering(s) from Vertex" }
             model_infos = offerings.map { |o| model_info_from_offering(o) }
             self.class.registry_publisher.publish_models_async(model_infos, readiness: readiness(live: false))
@@ -310,7 +320,7 @@ module Legion
 
           def offering_from_live_model(model)
             name = model['name'] || model['publisherModelName'] || model['model'] || model['id']
-            publisher = publisher_from_resource(name) || model['publisher'] || DEFAULT_PUBLISHER
+            publisher = publisher_from_resource(name) || model['publisher'] || default_publisher
             id = name.to_s.split('/').last
             offering_for(model: id, publisher:, metadata: model)
           end
@@ -320,8 +330,8 @@ module Legion
             Legion::Extensions::Llm::Routing::ModelOffering.new(
               provider_family: :vertex,
               instance_id: instance_id,
-              transport: configured_transport(:http),
-              tier: configured_tier(:frontier),
+              transport: offering_transport,
+              tier: offering_tier,
               model: model,
               usage_type: usage_type,
               capabilities: default_capabilities(model, api:),
@@ -337,16 +347,8 @@ module Legion
             )
           end
 
-          def configured_transport(default)
-            config.respond_to?(:transport) ? config.transport : default
-          end
-
-          def configured_tier(default)
-            config.respond_to?(:tier) ? config.tier : default
-          end
-
           def publisher_parent
-            "projects/#{project}/locations/#{location}/publishers/#{DEFAULT_PUBLISHER}/models"
+            "projects/#{project}/locations/#{location}/publishers/#{default_publisher}/models"
           end
 
           def publisher_model_path(model)
@@ -651,7 +653,7 @@ module Legion
             id = model_id(model)
             return publisher_from_resource(id) if id.start_with?('projects/')
 
-            PUBLISHERS.fetch(id, DEFAULT_PUBLISHER)
+            PUBLISHERS.fetch(id, default_publisher)
           end
 
           def publisher_from_resource(resource)
@@ -662,7 +664,7 @@ module Legion
           def api_for(model)
             id = model_id(model)
             return API_MODES[id] if API_MODES.key?(id)
-            return :raw_predict if publisher_for(id) != DEFAULT_PUBLISHER && !Capabilities.embeddings?(id)
+            return :raw_predict if publisher_for(id) != default_publisher && !Capabilities.embeddings?(id)
 
             :generate_content
           end

data/lib/legion/extensions/llm/vertex/version.rb

@@ -4,7 +4,7 @@ module Legion
   module Extensions
     module Llm
       module Vertex
-        VERSION = '0.2.8'
+        VERSION = '0.2.10'
       end
     end
   end

data/lib/legion/extensions/llm/vertex.rb

@@ -20,7 +20,10 @@ module Legion
             family: PROVIDER_FAMILY,
             instance: {
               endpoint: nil,
-              tier: :frontier,
+              project: nil,
+              location: 'us-central1',
+              publisher: 'google',
+              tier: :cloud,
               transport: :http,
               credentials: {
                 access_token: nil,
@@ -28,7 +31,7 @@ module Legion
               },
               provider: {
                 project: nil,
-                location: Provider::DEFAULT_LOCATION,
+                location: 'us-central1',
                 model_aliases: {}
               },
               usage: { inference: true, embedding: true, image: false },

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: lex-llm-vertex
 version: !ruby/object:Gem::Version
-  version: 0.2.8
+  version: 0.2.10
 platform: ruby
 authors:
 - LegionIO
@@ -98,6 +98,7 @@ files:
 - README.md
 - lex-llm-vertex.gemspec
 - lib/legion/extensions/llm/vertex.rb
+- lib/legion/extensions/llm/vertex/actors/discovery_refresh.rb
 - lib/legion/extensions/llm/vertex/actors/fleet_worker.rb
 - lib/legion/extensions/llm/vertex/provider.rb
 - lib/legion/extensions/llm/vertex/runners/fleet_worker.rb