flatten benchmark results in docs/ to a single menu item

docs/development/evaluations/history/.nav.yml

@@ -1,4 +1,5 @@
+sort:
+  direction: desc
 nav:
   - index.md
-  - Weekly: weekly/
-  - Special: special/
+  - "*"

docs/development/evaluations/history/index.md

@@ -1,18 +1,9 @@
 # Historical Evaluation Results
 
-## Weekly Runs
+Browse through our past benchmark runs to track performance trends over time.
 
-Weekly benchmark runs with a standard set of models.
+## Weekly Results
+Regular weekly benchmark runs that track model performance over time.
 
-See the **Weekly** section in the navigation sidebar for all weekly benchmark results.
-
-## Special Benchmark Runs
-
-One-off benchmark runs for specific purposes such as:
-
-- Comparing self-hosted models
-- Testing new model versions
-- Performance analysis for specific scenarios
-- Custom model comparisons
-
-See the **Special** section in the navigation sidebar for all special benchmark runs.
+## Extended Comparisons
+Special benchmark runs comparing multiple models and configurations.

run_benchmarks_local.sh

@@ -154,10 +154,10 @@ if [ -f "scripts/generate_eval_report.py" ]; then
         --models "$MODELS"
     echo "✅ Report generated: docs/development/evaluations/latest-results.md"
 
-    # Also generate timestamped version for history (always in weekly/)
-    mkdir -p docs/development/evaluations/history/weekly
+    # Also generate timestamped version for history
+    mkdir -p docs/development/evaluations/history
     TIMESTAMP=$(date +%Y%m%d_%H%M%S)
-    HISTORY_FILE="docs/development/evaluations/history/weekly/results_${TIMESTAMP}.md"
+    HISTORY_FILE="docs/development/evaluations/history/results_${TIMESTAMP}.md"
     poetry run python scripts/generate_eval_report.py \
         --json-file eval_results.json \
         --output-file "$HISTORY_FILE" \