+ lemans report: credit in summary

Author: palkan

CHANGELOG.md

@@ -1,6 +1,7 @@
 ## [Unreleased]
 
 - `lemant report` normalizes metrics (credit, cost, steps, and tokens) within each task individually before combining them in the final report.
+- `lemans report`: add credit to summary lines.
 
 ## [1.5.1] - 2026-10-06
 

lib/lemans/cli/report.rb

@@ -292,7 +292,19 @@ def stats(group)
         totals = self.class.tally(group).merge(cost_usd: group.sum { it[:cost_usd].to_f })
         rank = totals[:scored].positive? ? " (#{(100.0 * totals[:solved] / totals[:scored]).round}%)" : ""
         "#{totals[:total]} trials: #{totals[:scored]} scored, #{totals[:invalid]} invalid, " \
-          "#{totals[:solved]} solved#{rank} · $#{format("%.4f", totals[:cost_usd])}#{pass_at_k(group)}"
+          "#{totals[:solved]} solved#{rank}#{mean_credit(group)} · $#{format("%.4f", totals[:cost_usd])}" \
+          "#{pass_at_k(group)}"
+      end
+
+      # Each task weighs the same, as in the aggregate view
+      def mean_credit(group)
+        return "" unless fractional?
+
+        means = group.group_by { it[:task] }.values.filter_map do |rows|
+          credits = rows.filter_map { it[:credit] }
+          credits.sum(0.0) / credits.size unless credits.empty?
+        end
+        means.empty? ? "" : " · credit #{format("%.2f", means.sum / means.size)}"
       end
 
       def pass_at_k(group)

test/lemans/cli/report_test.rb

@@ -63,7 +63,7 @@ def test_the_rows_show_every_trial_and_the_summary_owns_up_to_the_totals
       assert_equal "-", invalid[rows.first.index("reward")]
       assert_equal "-", invalid[rows.first.index("credit")]
       assert_equal "-", invalid[rows.first.index("tokens")]
-      assert_includes report.summary_lines.join("\n"), "3 trials: 2 scored, 1 invalid, 1 solved (50%)"
+      assert_includes report.summary_lines.join("\n"), "3 trials: 2 scored, 1 invalid, 1 solved (50%) · credit 0.70"
     end
   end
 
@@ -76,6 +76,7 @@ def test_the_credit_column_appears_only_once_a_credit_differs_from_its_reward
 
       refute_includes report.to_rows.first, "credit"
       refute_includes Lemans::CLI::Report::Aggregate.new(report, keys: %i[task]).to_rows.first, "credit"
+      refute_includes report.summary_lines.join, "credit"
       assert_includes report.to_csv.lines.first, "credit"
     end
   end
@@ -115,16 +116,18 @@ def test_a_sweep_groups_the_summary_per_model_with_short_names
       store = Lemans::Stores::FS.new(runs_dir)
       store_result(store, trial: "a__1", model: "openrouter/openai/gpt-5.6-luna", reward: 1.0)
       store_result(store, trial: "a__2", model: "openrouter/openai/gpt-5.6-luna", reward: 0.0)
-      store_result(store, trial: "a__3", model: "openrouter/z-ai/glm-5.2", reward: 1.0)
+      store_result(store, trial: "a__3", model: "openrouter/z-ai/glm-5.2", reward: 0.0, credit: 0.6)
+      store_result(store, trial: "b__4", task: "b", model: "openrouter/z-ai/glm-5.2", reward: 1.0)
 
       report = Lemans::CLI::Report.load(store)
       lines = report.summary_lines
 
       assert_equal 3, lines.size
-      assert(lines[0..1].any? { it.start_with?("gpt-5.6-luna") && it.include?("1 solved (50%)") })
-      assert(lines[0..1].any? { it.start_with?("glm-5.2") && it.include?("1 solved (100%)") })
+      assert(lines[0..1].any? { it.start_with?("gpt-5.6-luna") && it.include?("1 solved (50%) · credit 0.50") })
+      assert(lines[0..1].any? { it.start_with?("glm-5.2") && it.include?("1 solved (50%) · credit 0.80") })
       assert lines.last.start_with?("total")
-      assert_includes lines.last, "2 solved (67%)"
+      # Each task weighs the same: a averages 0.53, b 1.0.
+      assert_includes lines.last, "2 solved (50%) · credit 0.77"
       # The table shows the short name; the CSV keeps the full provenance.
       assert_includes report.to_rows.flatten, "gpt-5.6-luna"
       assert_includes report.to_csv, "openrouter/openai/gpt-5.6-luna"