ruby_llm-agent-evaluations 0.2.0 → 0.2.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 58b4f6e5d15e47010525588cb3a6f57dd75b109aba1e6656a92f8b1d516632cb
-  data.tar.gz: 2422a756f17dc0719e86d56df65726a3162bfd66bf4e4cf11bff61b4b2a0387b
+  metadata.gz: 3c7cae718f0af572cfe99607d7d82e896ff635c332675c8e2df6c0e4b79f6a38
+  data.tar.gz: 43bae1d064264557c3c61293f7d1991ac951a43f7bec481e57371508e42f4726
 SHA512:
-  metadata.gz: 9533ace90d7b3fb21644ba6322826e7a8742cb070a7b0f281883e1184337cd1223be1a79d709f31144898cf3816511c9ab5106e2f00255193657b7b1281cf7eb
-  data.tar.gz: 2310dbd7d273e38c3812410db5b7660ec4843f21673bf9b8754040d0b9d86dcdcd57e3a1765fadebb67415393c325eaaeec6f84f61c9ce25aeb9eb9ad9aada15
+  metadata.gz: 070361d5bd53383114eb03ea872436af6e9501b4e19b6c0926671e948fce5e200ea7b151d4b12b15c79b8e87a6ddaf356bae858e8565c016c3e88ed826f2633e
+  data.tar.gz: a9124295de95c93f89e16565e7f73d6ede66a6210f59e7a517675ed9f9a2c4a640d298ea7cfdf9d11011b653a95240d7843a657896007b9485b5f847b3363967

data/CHANGELOG.md

@@ -1,5 +1,9 @@
 # Changelog
 
+## 0.2.1
+
+- Simplify new-project setup to one migration creating only runs, case results, and metric results, with `minimum_scores` from the start. Remove legacy schema detection, payload copying, temporary migration models, and the intermediate column-rename migration. Legacy database upgrades are no longer supported.
+
 ## 0.2.0
 
 - Publish this project as `ruby_llm-agent-evaluations` to avoid the separately owned RubyGems package. Rename the executable and gemspec; preserve `RubyLLM::Evaluations`, `ruby_llm/evaluations`, Rails migration tasks, and PostgreSQL history. Version 0.2.0 is the first RubyGems release under the new name; the earlier section records this project's development history.

data/README.md

@@ -146,9 +146,9 @@ bin/rails ruby_llm_evaluations:install:migrations
 bin/rails db:migrate
 ~~~
 
-When upgrading, repeat the install:migrations and db:migrate commands to install and apply new engine migrations; keep existing migrations and history. Standalone users instead run ruby_llm-agent-evaluations db:migrate before their next run.
+For standalone use, run `ruby_llm-agent-evaluations db:migrate` once before the first evaluation. Rails applications install the single engine migration and run `bin/rails db:migrate` as shown above.
 
-PostgreSQL is required. Tables are ruby_llm_evaluations_runs, ruby_llm_evaluations_case_results, and ruby_llm_evaluations_metric_results. Existing legacy payload history is migrated, not discarded. There are no Store objects, dynamically allocated models, or separate connection pools. Execution does not establish, migrate, or disconnect the host pool.
+PostgreSQL is required. The single migration creates only three evaluation tables: `ruby_llm_evaluations_runs`, `ruby_llm_evaluations_case_results`, and `ruby_llm_evaluations_metric_results`. It creates `minimum_scores` directly, with the foreign keys and indexes needed for history and baseline comparisons. No legacy tables, payload imports, or custom schema-version tables are required or inspected. ActiveRecord manages its ordinary migration bookkeeping. Execution uses the host connection pool and does not establish, migrate, or disconnect it.
 
 ~~~ruby
 # config/routes.rb
@@ -298,7 +298,7 @@ bundle exec rake integration
 bundle exec rake test
 
 gem build ruby_llm-agent-evaluations.gemspec
-gem install ./ruby_llm-agent-evaluations-0.2.0.gem
+gem install ./ruby_llm-agent-evaluations-0.2.1.gem
 ruby_llm-agent-evaluations --help
 ~~~
 

data/db/migrate/20261011000000_create_agent_evaluation_history.rb

@@ -4,44 +4,10 @@ class CreateAgentEvaluationHistory < ActiveRecord::Migration[7.1]
   RUNS = "ruby_llm_evaluations_runs"
   CASES = "ruby_llm_evaluations_case_results"
   METRICS = "ruby_llm_evaluations_metric_results"
-  LEGACY = "ruby_llm_evaluations_legacy_runs"
-  LEGACY_VERSIONS = "ruby_llm_evaluations_schema_versions"
 
-  class LegacyRun < ActiveRecord::Base
-    self.table_name = LEGACY
-    self.primary_key = "sequence"
-  end
-
-  class LegacyVersion < ActiveRecord::Base
-    self.table_name = LEGACY_VERSIONS
-    self.primary_key = "version"
-  end
-
-  class MigratedRun < ActiveRecord::Base
-    self.table_name = RUNS
-  end
-
-  class MigratedCase < ActiveRecord::Base
-    self.table_name = CASES
-  end
-
-  class MigratedMetric < ActiveRecord::Base
-    self.table_name = METRICS
-  end
-
-  def up
+  def change
     raise ActiveRecord::MigrationError, "evaluation history requires PostgreSQL" unless connection.adapter_name == "PostgreSQL"
 
-    legacy = table_exists?(RUNS) && column_exists?(RUNS, :payload)
-    if legacy
-      unless table_exists?(LEGACY_VERSIONS) && LegacyVersion.pluck(:version) == [1]
-        raise ActiveRecord::MigrationError, "unsupported legacy evaluation history version"
-      end
-      rename_table RUNS, LEGACY
-    elsif table_exists?(RUNS)
-      raise ActiveRecord::MigrationError, "evaluation model tables already exist without this migration"
-    end
-
     create_table RUNS, id: :string do |table|
       table.bigserial :sequence, null: false
       table.string :suite, null: false
@@ -68,7 +34,7 @@ class CreateAgentEvaluationHistory < ActiveRecord::Migration[7.1]
       table.jsonb :question
       table.jsonb :expected_answer
       table.jsonb :expected_actions
-      table.jsonb :criteria, null: false, default: {}
+      table.jsonb :minimum_scores, null: false, default: {}
       table.jsonb :metadata, null: false, default: {}
       table.jsonb :output
       table.string :status, null: false
@@ -97,49 +63,5 @@ class CreateAgentEvaluationHistory < ActiveRecord::Migration[7.1]
     end
     add_index METRICS, [:case_result_id, :position], unique: true, name: "rle_metric_order"
     add_index METRICS, [:case_result_id, :name], unique: true, name: "rle_metric_identity"
-
-    if legacy
-      copy_legacy_history
-      connection.reset_pk_sequence!(RUNS, "sequence", "#{RUNS}_sequence_seq")
-      drop_table LEGACY
-    end
-    drop_table LEGACY_VERSIONS if table_exists?(LEGACY_VERSIONS)
-  end
-
-  def down
-    raise ActiveRecord::IrreversibleMigration, "evaluation history migration preserves append-only evidence"
-  end
-
-  private
-
-  def copy_legacy_history
-    [LegacyRun, MigratedRun, MigratedCase, MigratedMetric].each(&:reset_column_information)
-    LegacyRun.find_each do |row|
-      value = row.payload
-      run = MigratedRun.create!(
-        id: value.fetch("id"), sequence: row.sequence, suite: value.fetch("suite"), agent: value.fetch("agent"),
-        version: value.fetch("version", "1"), suite_digest: value.fetch("suite_digest"),
-        started_at: value.fetch("started_at"), finished_at: value.fetch("finished_at"),
-        status: value.fetch("status"), metadata: value.fetch("metadata", {}), summary: value.fetch("summary")
-      )
-      value.fetch("cases").each_with_index do |item, position|
-        metrics = item.fetch("evaluations")
-        question = MigratedCase.create!(
-          run_id: run.id, position: position, case_id: item.fetch("id"), name: item.fetch("name"),
-          question: item["input"], expected_answer: item["expected"], expected_actions: item["expected_actions"],
-          criteria: metrics.to_h { |metric| [metric.fetch("name"), metric.fetch("threshold")] },
-          metadata: item.fetch("metadata", {}), output: item["output"], status: item.fetch("status"),
-          duration_ms: item.fetch("duration_ms"), input_tokens: item["input_tokens"], output_tokens: item["output_tokens"],
-          cost: item["cost"], actions: item["actions"], response_metadata: item["response_metadata"], error_details: item["error"]
-        )
-        metrics.each_with_index do |metric, metric_position|
-          MigratedMetric.create!(
-            case_result_id: question.id, position: metric_position, name: metric.fetch("name"),
-            threshold: metric.fetch("threshold"), score: metric["score"], passed: metric.fetch("passed"),
-            reason: metric["reason"], error_details: metric["error"]
-          )
-        end
-      end
-    end
   end
 end

data/lib/ruby_llm/evaluations/version.rb

@@ -2,6 +2,6 @@
 
 module RubyLLM
   module Evaluations
-    VERSION = "0.2.0"
+    VERSION = "0.2.1"
   end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: ruby_llm-agent-evaluations
 version: !ruby/object:Gem::Version
-  version: 0.2.0
+  version: 0.2.1
 platform: ruby
 authors:
 - RubyLLM Evaluations contributors
@@ -94,7 +94,6 @@ files:
 - app/models/ruby_llm/evaluations/run.rb
 - config/routes.rb
 - db/migrate/20261011000000_create_agent_evaluation_history.rb
-- db/migrate/20261011000100_rename_case_criteria_to_minimum_scores.rb
 - docs/images/dashboard-case.png
 - docs/images/dashboard-history.png
 - docs/images/dashboard-mobile.png

data/db/migrate/20261011000100_rename_case_criteria_to_minimum_scores.rb

@@ -1,7 +0,0 @@
-# frozen_string_literal: true
-
-class RenameCaseCriteriaToMinimumScores < ActiveRecord::Migration[7.1]
-  def change
-    rename_column :ruby_llm_evaluations_case_results, :criteria, :minimum_scores
-  end
-end