claude-evolve 1.5.3 → 1.5.4

package/bin/claude-evolve-analyze

@@ -513,6 +513,9 @@ print(f'max_desc=\"{desc_escaped}\"')
     # Calculate total data points for dynamic sizing
     total_data_points=$(awk 'END {print NR-1}' "$data_file")  # Subtract header row
     
+    # Count unique generations
+    unique_generations=$(awk '{if(NR>1) print $4}' "$data_file" | sort -nu | wc -l)
+    
     # AIDEV-NOTE: Dynamic dot sizing based on data point count
     # Use significantly larger dots when there are fewer data points for better visibility
     if [[ $total_data_points -lt 35 ]]; then
@@ -530,16 +533,24 @@ print(f'max_desc=\"{desc_escaped}\"')
     # Find all generations that have data
     generations=($(awk '{if(NR>1) print $4}' "$data_file" | sort -n | uniq))
     
-    for gen_num in "${generations[@]}"; do
-      if [[ -n $gen_num ]]; then
-        color=$(get_gen_color "$gen_num")
-        if [[ $gen_plots_added -gt 0 ]]; then
-          plot_cmd="$plot_cmd, \\"$'\n'
+    # If too many generations (>10), use a simplified plot without individual generation legends
+    if [[ $unique_generations -gt 10 ]]; then
+      # Single plot with color gradient based on generation number
+      plot_cmd="\"$data_file\" using 1:3:(\$4) with points palette pointsize $regular_dot_size notitle"
+      gen_plots_added=1
+    else
+      # Original plotting with individual generation legends
+      for gen_num in "${generations[@]}"; do
+        if [[ -n $gen_num ]]; then
+          color=$(get_gen_color "$gen_num")
+          if [[ $gen_plots_added -gt 0 ]]; then
+            plot_cmd="$plot_cmd, \\"$'\n'
+          fi
+          plot_cmd="${plot_cmd}     \"$data_file\" using (\$4==$gen_num?\$1:1/0):3 with points linecolor rgb \"$color\" pointsize $regular_dot_size title \"Gen $gen_num\""
+          ((gen_plots_added++))
         fi
-        plot_cmd="${plot_cmd}     \"$data_file\" using (\$4==$gen_num?\$1:1/0):3 with points linecolor rgb \"$color\" pointsize $regular_dot_size title \"Gen $gen_num\""
-        ((gen_plots_added++))
-      fi
-    done
+      done
+    fi
     
     # Add novel candidates
     if [[ -s "$novel_file" ]] && [[ $(wc -l < "$novel_file") -gt 1 ]]; then
@@ -586,6 +597,9 @@ print(f'max_desc=\"{desc_escaped}\"')
 set terminal png size 1200,800
 set output "$output_file"
 
+# Define unique generations count
+unique_gens = $unique_generations
+
 # Set up multiplot with proper spacing
 set multiplot layout 2,1 margins 0.08,0.82,0.15,0.95 spacing 0.1,0.15
 
@@ -595,7 +609,9 @@ set title "$EVOLUTION_CONTEXT Algorithm Evolution Performance Over Time" font ",
 unset xlabel
 set ylabel "Performance Score"
 set grid y  # Only show horizontal grid lines
-set key outside right
+
+# Show legend only if 10 or fewer generations
+if (unique_gens <= 10) set key outside right; else unset key
 
 # AIDEV-NOTE: Remove x-axis entirely to avoid tick problems with large datasets
 unset xtics
@@ -603,6 +619,9 @@ set autoscale
 set yrange [*:*]  # Auto-scale y-axis only
 
 # Define colors for generations
+# Use palette for many generations
+if (unique_gens > 10) set palette model RGB defined (0 "#1f77b4", 1 "#ff7f0e", 2 "#2ca02c", 3 "#d62728", 4 "#9467bd", 5 "#8c564b", 6 "#e377c2")
+
 plot $plot_cmd
 
 #=================== BOTTOM PLOT: Generation Medians ===================
@@ -614,8 +633,13 @@ set boxwidth 0.6
 unset key
 set grid y
 
-# Set custom x-axis labels
-set xtics ($xtics_labels)
+# Set custom x-axis labels (but hide if too many generations)
+if (unique_gens > 10) {
+    set xtics auto
+    set xtics rotate by -45
+} else {
+    set xtics ($xtics_labels)
+}
 
 # Auto-scale for generation plot too
 set autoscale

package/bin/claude-evolve-edit

@@ -20,7 +20,10 @@ show_help() {
 claude-evolve edit - Manage evolution candidate statuses by generation or status
 
 USAGE:
-  claude-evolve edit <selector> <action>
+  claude-evolve edit [--recent-generations=N] <selector> <action>
+
+OPTIONS:
+  --recent-generations=N  Limit operations to the most recent N generations only
 
 SELECTORS:
   gen01, gen02, etc.  Target specific generation
@@ -41,29 +44,56 @@ ACTIONS:
   delete           Delete candidates from CSV and remove .py files (asks confirmation)
 
 EXAMPLES:
-  claude-evolve edit gen03 failed        # Mark all gen03 as failed
-  claude-evolve edit failed pending      # Reset all failed candidates to pending
-  claude-evolve edit failed failed-retry1 # Convert failed to retry status (bug fixing)
-  claude-evolve edit complete failed     # Mark all complete as failed for re-run
-  claude-evolve edit all pending         # Mark everything as pending for re-run
-  claude-evolve edit gen02 reboot        # Full reset of gen02 (delete files + clear data)
-  claude-evolve edit gen02 delete        # Delete gen02 from CSV and remove .py files
+  claude-evolve edit gen03 failed                    # Mark all gen03 as failed
+  claude-evolve edit failed pending                  # Reset all failed candidates to pending
+  claude-evolve edit --recent-generations=15 failed pending  # Reset only recent 15 gen failures
+  claude-evolve edit --recent-generations=5 complete failed  # Re-run recent 5 gen completions
+  claude-evolve edit failed failed-retry1            # Convert failed to retry status (bug fixing)
+  claude-evolve edit complete failed                 # Mark all complete as failed for re-run
+  claude-evolve edit all pending                     # Mark everything as pending for re-run
+  claude-evolve edit gen02 reboot                    # Full reset of gen02 (delete files + clear data)
+  claude-evolve edit gen02 delete                    # Delete gen02 from CSV and remove .py files
 
 DESCRIPTION:
   This command helps manage evolution runs when you need to re-evaluate candidates.
   Use status selectors (failed, complete, etc.) to bulk-change candidates by status.
   Use 'reboot' for complete reset including file deletion.
+  Use --recent-generations to limit operations to recent work only, useful for large systems.
 EOF
 }
 
 # Parse arguments
-if [[ $# -ne 2 ]]; then
+recent_generations=""
+args=()
+
+while [[ $# -gt 0 ]]; do
+  case "$1" in
+    --recent-generations=*)
+      recent_generations="${1#*=}"
+      if [[ ! "$recent_generations" =~ ^[1-9][0-9]*$ ]]; then
+        echo "[ERROR] --recent-generations must be a positive integer" >&2
+        exit 1
+      fi
+      shift
+      ;;
+    --help|-h)
+      show_help
+      exit 0
+      ;;
+    *)
+      args+=("$1")
+      shift
+      ;;
+  esac
+done
+
+if [[ ${#args[@]} -ne 2 ]]; then
   show_help
   exit 1
 fi
 
-SELECTOR="$1"
-ACTION="$2"
+SELECTOR="${args[0]}"
+ACTION="${args[1]}"
 
 # Validate configuration
 if ! validate_config; then
@@ -99,7 +129,11 @@ update_candidates_status() {
   local new_status="$2"
   local clear_scores="$3"
   
-  echo "[INFO] Updating candidates matching '$selector' to status: $new_status"
+  local filter_msg=""
+  if [[ -n "$recent_generations" ]]; then
+    filter_msg=" (limited to recent $recent_generations generations)"
+  fi
+  echo "[INFO] Updating candidates matching '$selector' to status: $new_status${filter_msg}"
   
   # Use Python to safely edit the CSV
   "$PYTHON_CMD" -c "
@@ -112,6 +146,7 @@ csv_file = '$FULL_CSV_PATH'
 selector = '$selector'
 new_status = '$new_status'
 clear_scores = '$clear_scores' == 'true'
+recent_generations = '$recent_generations'
 
 
 try:
@@ -127,6 +162,31 @@ try:
     header = rows[0]
     updated_count = 0
     
+    # If recent_generations is specified, determine which generations to include
+    recent_gen_set = set()
+    if recent_generations and recent_generations.isdigit():
+        n_recent = int(recent_generations)
+        
+        # Find all generation numbers from candidate IDs
+        all_generations = set()
+        for i in range(1, len(rows)):
+            row = rows[i]
+            if len(row) < 1:
+                continue
+            candidate_id = row[0]
+            
+            # Extract generation number from candidate_id (e.g., gen01-001 -> 1)
+            match = re.match(r'^gen(\d+)-', candidate_id)
+            if match:
+                gen_num = int(match.group(1))
+                all_generations.add(gen_num)
+        
+        # Get the most recent N generations
+        if all_generations:
+            sorted_generations = sorted(all_generations, reverse=True)
+            recent_gen_set = set(sorted_generations[:n_recent])
+            print(f'[INFO] Filtering to recent generations: {sorted(recent_gen_set)}', file=sys.stderr)
+    
     # Update matching rows
     for i in range(1, len(rows)):
         row = rows[i]
@@ -152,6 +212,18 @@ try:
             else:
                 matches = current_status == selector
         
+        # Apply recent generations filter if specified
+        if matches and recent_gen_set:
+            # Extract generation number from candidate_id
+            gen_match = re.match(r'^gen(\d+)-', candidate_id)
+            if gen_match:
+                candidate_gen = int(gen_match.group(1))
+                if candidate_gen not in recent_gen_set:
+                    matches = False  # Filter out this candidate
+            else:
+                # Non-generation candidates (like baseline) - exclude when filtering by recent generations
+                matches = False
+        
         if matches:
             if clear_scores:
                 # Clear everything after description (keep id, basedOnId, description)
@@ -192,6 +264,12 @@ delete_evolution_files() {
     return
   fi
   
+  local filter_msg=""
+  if [[ -n "$recent_generations" ]]; then
+    filter_msg=" (limited to recent $recent_generations generations)"
+  fi
+  echo "[INFO] Deleting evolution files for '$selector'${filter_msg}..."
+  
   local deleted_count=0
   
   if [[ "$selector" == "all" ]]; then
@@ -225,15 +303,46 @@ import re
 
 csv_file = '$FULL_CSV_PATH'
 selector = '$selector'
+recent_generations = '$recent_generations'
 
 
 try:
     with open(csv_file, 'r') as f:
         reader = csv.reader(f)
-        next(reader)  # Skip header
+        rows = list(reader)
+        
+        if not rows:
+            print('')
+            sys.exit(0)
+        
+        # Skip header if present
+        start_idx = 1 if rows and rows[0] and rows[0][0].lower() == 'id' else 0
+        
+        # Determine recent generations if filtering is requested
+        recent_gen_set = set()
+        if recent_generations and recent_generations.isdigit():
+            n_recent = int(recent_generations)
+            
+            # Find all generation numbers from candidate IDs
+            all_generations = set()
+            for row in rows[start_idx:]:
+                if len(row) < 1:
+                    continue
+                candidate_id = row[0]
+                
+                # Extract generation number from candidate_id (e.g., gen01-001 -> 1)
+                match = re.match(r'^gen(\d+)-', candidate_id)
+                if match:
+                    gen_num = int(match.group(1))
+                    all_generations.add(gen_num)
+            
+            # Get the most recent N generations
+            if all_generations:
+                sorted_generations = sorted(all_generations, reverse=True)
+                recent_gen_set = set(sorted_generations[:n_recent])
         
         candidates = []
-        for row in reader:
+        for row in rows[start_idx:]:
             if len(row) < 1:
                 continue
             
@@ -249,6 +358,18 @@ try:
             else:
                 matches = current_status == selector
             
+            # Apply recent generations filter if specified
+            if matches and recent_gen_set:
+                # Extract generation number from candidate_id
+                gen_match = re.match(r'^gen(\d+)-', candidate_id)
+                if gen_match:
+                    candidate_gen = int(gen_match.group(1))
+                    if candidate_gen not in recent_gen_set:
+                        matches = False  # Filter out this candidate
+                else:
+                    # Non-generation candidates (like baseline) - exclude when filtering by recent generations
+                    matches = False
+            
             if matches:
                 candidates.append(candidate_id)
         
@@ -284,7 +405,11 @@ except Exception as e:
 delete_candidates_from_csv() {
   local selector="$1"
   
-  echo "[INFO] Deleting candidates matching '$selector' from CSV..."
+  local filter_msg=""
+  if [[ -n "$recent_generations" ]]; then
+    filter_msg=" (limited to recent $recent_generations generations)"
+  fi
+  echo "[INFO] Deleting candidates matching '$selector' from CSV${filter_msg}..."
   
   "$PYTHON_CMD" -c "
 import sys
@@ -293,6 +418,7 @@ from lib.evolution_csv import EvolutionCSV
 import re
 
 selector = '$selector'
+recent_generations = '$recent_generations'
 deleted_count = 0
 
 with EvolutionCSV('$FULL_CSV_PATH') as csv:
@@ -306,6 +432,29 @@ with EvolutionCSV('$FULL_CSV_PATH') as csv:
         has_header = rows and rows[0] and rows[0][0].lower() == 'id'
         start_idx = 1 if has_header else 0
         
+        # Determine recent generations if filtering is requested
+        recent_gen_set = set()
+        if recent_generations and recent_generations.isdigit():
+            n_recent = int(recent_generations)
+            
+            # Find all generation numbers from candidate IDs
+            all_generations = set()
+            for row in rows[start_idx:]:
+                if not row or not row[0].strip():
+                    continue
+                candidate_id = row[0].strip()
+                
+                # Extract generation number from candidate_id (e.g., gen01-001 -> 1)
+                match = re.match(r'^gen(\d+)-', candidate_id)
+                if match:
+                    gen_num = int(match.group(1))
+                    all_generations.add(gen_num)
+            
+            # Get the most recent N generations
+            if all_generations:
+                sorted_generations = sorted(all_generations, reverse=True)
+                recent_gen_set = set(sorted_generations[:n_recent])
+        
         for row in rows[start_idx:]:
             if not row or not row[0].strip():
                 continue
@@ -327,6 +476,18 @@ with EvolutionCSV('$FULL_CSV_PATH') as csv:
             else:
                 matches = current_status == selector
             
+            # Apply recent generations filter if specified
+            if matches and recent_gen_set:
+                # Extract generation number from candidate_id
+                gen_match = re.match(r'^gen(\d+)-', candidate_id)
+                if gen_match:
+                    candidate_gen = int(gen_match.group(1))
+                    if candidate_gen not in recent_gen_set:
+                        matches = False  # Filter out this candidate
+                else:
+                    # Non-generation candidates (like baseline) - exclude when filtering by recent generations
+                    matches = False
+            
             if matches:
                 candidates_to_delete.append(candidate_id)
     
@@ -341,7 +502,11 @@ with EvolutionCSV('$FULL_CSV_PATH') as csv:
 }
 
 # Main execution
-echo "[INFO] Processing '$SELECTOR' with action: $ACTION"
+info_msg="Processing '$SELECTOR' with action: $ACTION"
+if [[ -n "$recent_generations" ]]; then
+  info_msg="$info_msg (limited to recent $recent_generations generations)"
+fi
+echo "[INFO] $info_msg"
 
 case "$ACTION" in
   failed)
@@ -387,4 +552,4 @@ echo "[INFO] Edit operation complete"
 
 # Call status command to show current state
 echo ""
-"$SCRIPT_DIR/claude-evolve-status" --brief
+"$SCRIPT_DIR/claude-evolve-status" --brief

package/bin/claude-evolve-ideate

@@ -84,32 +84,77 @@ call_ai_for_ideation() {
     original_csv_count=0
   fi
   
-  # Use centralized AI library
-  local ai_output
-  ai_output=$(call_ai_with_round_robin "$prompt" "ideate" "$gen_num")
-  local ai_exit_code=$?
+  echo "[DEBUG] Original CSV has $original_csv_count data rows" >&2
   
-  # Handle special exit codes
-  # Always check if the file was modified, regardless of exit code
-  # This is the only test that matters - did the CSV change?
+  # Get models for ideation
+  local model_list
+  model_list=$(get_models_for_command "ideate")
+  local models=()
+  read -ra models <<< "$model_list"
   
-  if [[ -f "$temp_csv_file" ]]; then
-    local new_csv_count
-    new_csv_count=$(grep -v '^[[:space:]]*$' "$temp_csv_file" | tail -n +2 | wc -l)
+  if [[ ${#models[@]} -eq 0 ]]; then
+    echo "[ERROR] No models configured for ideation" >&2
+    return 1
+  fi
+  
+  # Calculate starting index for round-robin
+  local num_models=${#models[@]}
+  local start_index=$((gen_num % num_models))
+  
+  # Create ordered list based on round-robin
+  local ordered_models=()
+  for ((i=0; i<num_models; i++)); do
+    local idx=$(((start_index + i) % num_models))
+    ordered_models+=("${models[$idx]}")
+  done
+  
+  echo "[AI] Model order for ideate (round-robin): ${ordered_models[*]}" >&2
+  
+  # Try each model until CSV changes
+  for model in "${ordered_models[@]}"; do
+    echo "[AI] Attempting ideate with $model" >&2
+    
+    # Call the model directly
+    local ai_output
+    ai_output=$(call_ai_model_configured "$model" "$prompt")
+    local ai_exit_code=$?
+    
+    echo "[AI] $model completed with exit code $ai_exit_code" >&2
     
-    if [[ $new_csv_count -gt $original_csv_count ]]; then
-      echo "[INFO] CSV was modified ($new_csv_count vs $original_csv_count rows) - success!" >&2
-      return 0
+    # Check if the file was modified - this is ALL that matters
+    if [[ -f "$temp_csv_file" ]]; then
+      local new_csv_count
+      new_csv_count=$(grep -v '^[[:space:]]*$' "$temp_csv_file" | tail -n +2 | wc -l)
+      
+      if [[ $new_csv_count -gt $original_csv_count ]]; then
+        echo "[INFO] CSV was modified by $model ($new_csv_count vs $original_csv_count rows) - validating format..." >&2
+        
+        # Post-process to ensure all description fields are quoted
+        local fixed_csv_file="${temp_csv_file}.fixed"
+        
+        # Use the CSV fixer script
+        if "$PYTHON_CMD" "$SCRIPT_DIR/../lib/csv_fixer.py" "$temp_csv_file" "$fixed_csv_file"; then
+          mv "$fixed_csv_file" "$temp_csv_file"
+          echo "[INFO] CSV format validated and fixed if needed" >&2
+        else
+          echo "[WARN] CSV format validation failed, using original" >&2
+        fi
+        
+        return 0
+      else
+        echo "[INFO] CSV unchanged after $model (exit code: $ai_exit_code)" >&2
+        echo "[DEBUG] Original count: $original_csv_count, New count: $new_csv_count" >&2
+        # Continue to next model
+      fi
     else
-      echo "[INFO] CSV unchanged after AI call (exit code: $ai_exit_code)" >&2
-      echo "[DEBUG] Original count: $original_csv_count, New count: $new_csv_count" >&2
-      return 1
+      echo "[INFO] Temp CSV file not found after $model: $temp_csv_file" >&2
+      # Continue to next model
     fi
-  else
-    echo "[INFO] Temp CSV file not found after AI call: $temp_csv_file" >&2
-    echo "[DEBUG] Exit code was: $ai_exit_code" >&2
-    return 1
-  fi
+  done
+  
+  # All models tried, none changed the file
+  echo "[ERROR] All AI models failed to generate ideas" >&2
+  return 1
 }
 
 # Parse arguments
@@ -455,8 +500,8 @@ Instructions:
 3. For each idea, create a row with: id,parent_id,description,,pending
 4. CRITICAL CSV FORMATTING RULES:
    - ALWAYS wrap the description field in double quotes
-   - If the description contains quotes, escape them by doubling them (" becomes "")
-   - Example: gen01-001,gen00-000,"Implement adaptive RSI thresholds",,pending
+   - If the description contains quotes, escape them by doubling them (\" becomes \"\")
+   - Example: gen01-001,gen00-000,\"Implement adaptive RSI thresholds\",,pending
    - BAD: gen01-001,gen00-000,Implement adaptive RSI thresholds,,pending
    - NEVER omit quotes - unquoted descriptions cause CSV corruption
 5. For novel ideas: leave parent_id empty
@@ -839,8 +884,8 @@ CRITICAL INSTRUCTIONS:
 7. For each idea, create a row with: id,,description,,pending (empty parent_id for novel ideas)
 8. CRITICAL CSV FORMATTING RULES:
    - ALWAYS wrap the description field in double quotes
-   - If the description contains quotes, escape them by doubling them (" becomes "")
-   - Example: gen01-001,,"Implement adaptive RSI thresholds based on volatility",,pending
+   - If the description contains quotes, escape them by doubling them (\" becomes \"\")
+   - Example: gen01-001,,\"Implement adaptive RSI thresholds based on volatility\",,pending
    - BAD: gen01-001,,Implement adaptive RSI thresholds based on volatility,,pending
    - NEVER omit quotes around descriptions - this causes CSV parsing errors
 9. Each description should be one clear sentence describing a novel algorithmic approach
@@ -945,12 +990,12 @@ CRITICAL INSTRUCTIONS:
 7. Each parent_id MUST be one of: $valid_parent_ids
 8. CRITICAL CSV FORMATTING RULES:
    - ALWAYS wrap the description field in double quotes
-   - If the description contains quotes, escape them by doubling them (" becomes "")
-   - Example: gen01-001,gen00-000,"Lower rsi_entry from 21 to 18",,pending
+   - If the description contains quotes, escape them by doubling them (\" becomes \"\")
+   - Example: gen01-001,gen00-000,\"Lower rsi_entry from 21 to 18\",,pending
    - BAD: gen01-001,gen00-000,Lower rsi_entry from 21 to 18,,pending
    - NEVER omit quotes around descriptions - this causes CSV parsing errors
 9. Each description should focus on adjusting specific parameters that exist in the parent's source code
-10. Include current and new parameter values (e.g., "Lower rsi_entry from 21 to 18")
+10. Include current and new parameter values - for example: \"Lower rsi_entry from 21 to 18\"
 
 IMPORTANT: You must APPEND new rows to the existing CSV file. DO NOT replace the file contents. All existing rows must remain unchanged.
 CRITICAL: You must use your file editing tools (Edit/MultiEdit) to modify the CSV file. DO NOT return CSV text - use your tools to edit the file directly.
@@ -1032,8 +1077,8 @@ CRITICAL INSTRUCTIONS:
 7. Each parent_id MUST be one of: $valid_parent_ids
 8. CRITICAL CSV FORMATTING RULES:
    - ALWAYS wrap the description field in double quotes
-   - If the description contains quotes, escape them by doubling them (" becomes "")
-   - Example: gen01-001,gen00-000,"Add ML-based regime detection using LSTM",,pending
+   - If the description contains quotes, escape them by doubling them (\" becomes \"\")
+   - Example: gen01-001,gen00-000,\"Add ML-based regime detection using LSTM\",,pending
    - BAD: gen01-001,gen00-000,Add ML-based regime detection using LSTM,,pending
    - NEVER omit quotes around descriptions - this causes CSV parsing errors
 9. Each description should focus on architectural/structural changes based on the parent's actual code
@@ -1119,8 +1164,8 @@ CRITICAL INSTRUCTIONS:
 7. Each parent_id MUST be one of: $valid_parent_ids (choose the primary parent)
 8. CRITICAL CSV FORMATTING RULES:
    - ALWAYS wrap the description field in double quotes
-   - If the description contains quotes, escape them by doubling them (" becomes "")
-   - Example: gen01-001,gen00-000,"Combine gen01-123's RSI logic with gen01-456's volatility scaling",,pending
+   - If the description contains quotes, escape them by doubling them (\" becomes \"\")
+   - Example: gen01-001,gen00-000,\"Combine gen01-123's RSI logic with gen01-456's volatility scaling\",,pending
    - BAD: gen01-001,gen00-000,Combine gen01-123's RSI logic with gen01-456's volatility scaling,,pending
    - NEVER omit quotes around descriptions - this causes CSV parsing errors
 9. Each description should combine actual elements from 2+ algorithms based on their source code
@@ -1219,8 +1264,8 @@ CRITICAL INSTRUCTIONS:
 6. For each idea, create a row with: id,parent_id,description,,pending
 7. CRITICAL CSV FORMATTING RULES:
    - ALWAYS wrap the description field in double quotes
-   - If the description contains quotes, escape them by doubling them (" becomes "")
-   - Example: gen01-001,gen00-000,"Implement adaptive RSI thresholds based on volatility",,pending
+   - If the description contains quotes, escape them by doubling them (\" becomes \"\")
+   - Example: gen01-001,gen00-000,\"Implement adaptive RSI thresholds based on volatility\",,pending
    - BAD: gen01-001,gen00-000,Implement adaptive RSI thresholds based on volatility,,pending
    - NEVER omit quotes around descriptions - this causes CSV parsing errors that corrupt the data
 8. Mix both parameter tuning and structural changes

package/bin/claude-evolve-run

@@ -478,6 +478,66 @@ ensure_baseline_entry
 # Flag to track API limit status
 api_limit_reached=false
 
+# Check if previous generation has at least one completed item
+check_previous_generation_has_completed() {
+  local csv_path="$1"
+  
+  if [[ ! -f "$csv_path" ]]; then
+    # No CSV file yet - allow first ideation
+    return 0
+  fi
+  
+  # Use Python to find the maximum generation and check if it has completed items
+  "$PYTHON_CMD" -c "
+import csv
+import sys
+
+max_gen = 0
+gen_completed = {}
+
+try:
+    with open('$csv_path', 'r') as f:
+        reader = csv.reader(f)
+        next(reader, None)  # Skip header
+        for row in reader:
+            if row and len(row) >= 5:
+                id_field = row[0].strip()
+                status = row[4].strip() if len(row) > 4 else 'pending'
+                
+                if id_field.startswith('gen') and '-' in id_field:
+                    try:
+                        gen_part = id_field.split('-')[0]  # e.g., 'gen01'
+                        gen_num = int(gen_part[3:])  # Extract number after 'gen'
+                        max_gen = max(max_gen, gen_num)
+                        
+                        if gen_num not in gen_completed:
+                            gen_completed[gen_num] = 0
+                        
+                        if status == 'complete':
+                            gen_completed[gen_num] += 1
+                    except (ValueError, IndexError):
+                        pass
+
+    # If max_gen is 0, no generations exist yet - allow ideation
+    if max_gen == 0:
+        sys.exit(0)
+    
+    # Check if the most recent generation has at least one completed item
+    if gen_completed.get(max_gen, 0) > 0:
+        print(f'[INFO] Generation {max_gen:02d} has {gen_completed[max_gen]} completed items - allowing ideation', file=sys.stderr)
+        sys.exit(0)
+    else:
+        print(f'[INFO] Generation {max_gen:02d} has no completed items - blocking ideation to prevent endless loops', file=sys.stderr)
+        sys.exit(1)
+
+except Exception as e:
+    print(f'[ERROR] Failed to check previous generation: {e}', file=sys.stderr)
+    # On error, allow ideation to maintain backward compatibility
+    sys.exit(0)
+"
+  return $?
+}
+
 # Main dispatch loop
 while true; do
   # Clean up finished workers
@@ -506,7 +566,17 @@ while true; do
     
     # Check if auto ideation is enabled
     if [[ "$AUTO_IDEATE" == "true" || "$AUTO_IDEATE" == "1" ]]; then
-      echo "[DISPATCHER] Auto ideation is enabled. Generating new ideas..."
+      echo "[DISPATCHER] Auto ideation is enabled. Checking prerequisites..."
+      
+      # Check if previous generation has at least one completed item
+      if ! check_previous_generation_has_completed "$FULL_CSV_PATH"; then
+        echo "[DISPATCHER] Evolution complete - previous generation has no completed items."
+        echo "[DISPATCHER] This prevents endless ideation loops when API limits are hit."
+        echo "[DISPATCHER] Wait for current generation to complete, then run 'claude-evolve ideate' manually."
+        break
+      fi
+      
+      echo "[DISPATCHER] Prerequisites met. Generating new ideas..."
       
       # Check if claude-evolve-ideate exists
       ideate_script="$SCRIPT_DIR/claude-evolve-ideate"

package/bin/claude-evolve-worker

@@ -14,34 +14,11 @@ TERMINATION_SIGNAL=""
 # Cleanup function to handle termination
 cleanup_on_exit() {
   if [[ -n "$CURRENT_CANDIDATE_ID" ]]; then
-    # Only mark as failed if it was a timeout (SIGTERM from timeout command)
-    # For user interruption (Ctrl-C) or kill, leave it for retry
-    if [[ "$TERMINATION_SIGNAL" == "TERM" ]]; then
-      echo "[WORKER-$$] Timeout detected, marking $CURRENT_CANDIDATE_ID as failed" >&2
-      "$PYTHON_CMD" -c "
-import sys
-sys.path.insert(0, '$SCRIPT_DIR/..')
-from lib.evolution_csv import EvolutionCSV
-try:
-    with EvolutionCSV('$FULL_CSV_PATH') as csv:
-        csv.update_candidate_status('$CURRENT_CANDIDATE_ID', 'failed')
-except:
-    pass  # Best effort cleanup
-" 2>/dev/null || true
-    else
-      echo "[WORKER-$$] Interrupted, leaving $CURRENT_CANDIDATE_ID for retry" >&2
-      # Optionally reset to pending instead of leaving as running
-      "$PYTHON_CMD" -c "
-import sys
-sys.path.insert(0, '$SCRIPT_DIR/..')
-from lib.evolution_csv import EvolutionCSV
-try:
-    with EvolutionCSV('$FULL_CSV_PATH') as csv:
-        csv.update_candidate_status('$CURRENT_CANDIDATE_ID', 'pending')
-except:
-    pass  # Best effort cleanup
-" 2>/dev/null || true
-    fi
+    echo "[WORKER-$$] Worker terminated while processing $CURRENT_CANDIDATE_ID" >&2
+    # If we're interrupted while processing, leave it as "running"
+    # This prevents other workers from picking it up in the same session
+    # A human can manually reset to pending if needed
+    echo "[WORKER-$$] Leaving $CURRENT_CANDIDATE_ID in current state" >&2
   fi
 }
 
@@ -135,6 +112,12 @@ process_candidate() {
   echo "[WORKER-$$] Description: $description"
   echo "[WORKER-$$] Based on ID: $parent_id"
   
+  # Treat "baseline-000" parent ID as empty/baseline
+  if [[ "$parent_id" == "baseline-000" ]]; then
+    parent_id=""
+    echo "[WORKER-$$] Parent ID 'baseline-000' treated as baseline (empty parent)"
+  fi
+  
   # Determine source algorithm
   local source_file
   if [[ -z "$parent_id" ]]; then
@@ -213,16 +196,26 @@ CRITICAL: Do NOT use any git commands (git add, git commit, git reset, etc.). On
       
       # Try AI models with round-robin based on candidate ID
       if ! call_ai_for_evolution "$evolution_prompt" "$candidate_id"; then
-        echo "[WORKER-$$] ERROR: All AI models failed to generate code" >&2
+        echo "[WORKER-$$] ERROR: All AI models failed to generate code - leaving as pending for retry" >&2
         cd "$original_pwd"
         rm -f "$target_file"  # Clean up on failure
-        return 1
+        # Return with special code to indicate AI failure (should remain pending)
+        return 77
       fi
       
       # Restore working directory
       cd "$original_pwd"
       
       echo "[WORKER-$$] Evolution applied successfully"
+      
+      # Check if the generated Python file has syntax errors
+      echo "[WORKER-$$] Checking Python syntax..." >&2
+      if ! "$PYTHON_CMD" -m py_compile "$target_file" 2>&1; then
+        echo "[WORKER-$$] ERROR: Generated Python file has syntax errors!" >&2
+        echo "[WORKER-$$] File: $target_file" >&2
+        # This is still an evaluation failure, not an AI failure
+        return 1
+      fi
     fi
   fi
   
@@ -239,11 +232,22 @@ CRITICAL: Do NOT use any git commands (git add, git commit, git reset, etc.). On
     eval_arg=""
   fi
   local eval_cmd=("$PYTHON_CMD" "$FULL_EVALUATOR_PATH" "$eval_arg")
+  
+  # Add memory limiting if configured
+  if [[ -n "$MEMORY_LIMIT_MB" ]] && [[ "$MEMORY_LIMIT_MB" -gt 0 ]]; then
+    eval_cmd=("$PYTHON_CMD" "$SCRIPT_DIR/../lib/memory_limit_wrapper.py" "$MEMORY_LIMIT_MB" "${eval_cmd[@]}")
+  fi
+  
+  # Add timeout if configured
   [[ -n "$timeout_seconds" ]] && eval_cmd=(timeout "$timeout_seconds" "${eval_cmd[@]}")
   
   # Run evaluation with tee to both display and capture output
   # Use stdbuf to disable buffering for real-time output
-  if stdbuf -o0 -e0 "${eval_cmd[@]}" 2>&1 | tee "$eval_output_file" >&2; then
+  # IMPORTANT: Use PIPESTATUS to get the exit code of the evaluation command, not tee
+  stdbuf -o0 -e0 "${eval_cmd[@]}" 2>&1 | tee "$eval_output_file" >&2
+  local eval_exit_code=${PIPESTATUS[0]}  # Get exit code of first command in pipe
+  
+  if [[ $eval_exit_code -eq 0 ]]; then
     local eval_end=$(date +%s)
     local eval_duration=$((eval_end - eval_start))
     
@@ -353,14 +357,19 @@ with EvolutionCSV('$FULL_CSV_PATH') as csv:
       echo "[WORKER-$$] Output: $eval_output" >&2
       # rm -f "$eval_output_file"  # Keep for debugging
       echo "[WORKER-$$] Evaluation output saved to: $eval_output_file" >&2
+      # Clear CURRENT_CANDIDATE_ID before returning
+      CURRENT_CANDIDATE_ID=""
       return 1
     fi
     
     # Clean up temp file (comment out to keep for debugging)
     # rm -f "$eval_output_file"
     echo "[WORKER-$$] Evaluation output saved to: $eval_output_file" >&2
+    
+    # Clear CURRENT_CANDIDATE_ID on successful completion
+    CURRENT_CANDIDATE_ID=""
   else
-    local exit_code=$?
+    local exit_code=$eval_exit_code
     # Read any output that was captured before failure
     eval_output=$(<"$eval_output_file")
     # rm -f "$eval_output_file"  # Keep for debugging
@@ -370,22 +379,56 @@ with EvolutionCSV('$FULL_CSV_PATH') as csv:
     echo "[WORKER-$$] Output: $eval_output" >&2
     
     # Mark as failed in CSV
-    "$PYTHON_CMD" -c "
+    echo "[WORKER-$$] Marking $candidate_id as failed in CSV" >&2
+    if ! "$PYTHON_CMD" -c "
 import sys
 sys.path.insert(0, '$SCRIPT_DIR/..')
 from lib.evolution_csv import EvolutionCSV
-with EvolutionCSV('$FULL_CSV_PATH') as csv:
-    csv.update_candidate_status('$candidate_id', 'failed')
-"
+try:
+    with EvolutionCSV('$FULL_CSV_PATH') as csv:
+        success = csv.update_candidate_status('$candidate_id', 'failed')
+        if not success:
+            print(f'ERROR: Failed to update status for {candidate_id}', file=sys.stderr)
+            sys.exit(1)
+except Exception as e:
+    print(f'ERROR: Exception updating status: {e}', file=sys.stderr)
+    sys.exit(1)
+" 2>&1; then
+      echo "[WORKER-$$] ERROR: Failed to update CSV status to failed" >&2
+    else
+      echo "[WORKER-$$] Successfully marked $candidate_id as failed" >&2
+    fi
     
+    # Clear CURRENT_CANDIDATE_ID before returning to prevent cleanup handler from resetting it
+    CURRENT_CANDIDATE_ID=""
     return $exit_code
   fi
 }
 
+# Don't reset running candidates on startup - they might be legitimately being processed by another worker
+
 # Main worker loop
 echo "[WORKER-$$] Worker started"
 
 while true; do
+  # Debug: Show current status of all candidates
+  echo "[WORKER-$$] Current candidate statuses:" >&2
+  "$PYTHON_CMD" -c "
+import sys
+sys.path.insert(0, '$SCRIPT_DIR/..')
+from lib.evolution_csv import EvolutionCSV
+with EvolutionCSV('$FULL_CSV_PATH') as csv:
+    rows = csv._read_csv()
+    if rows:
+        start_idx = 1 if rows and rows[0] and rows[0][0].lower() == 'id' else 0
+        status_count = {}
+        for row in rows[start_idx:]:
+            if len(row) > 4:
+                status = row[4].strip() or 'pending'
+                status_count[status] = status_count.get(status, 0) + 1
+        print(f'Status counts: {status_count}', file=sys.stderr)
+" 2>&1 || true
+
   # Try to claim a pending candidate
   candidate_info=$("$PYTHON_CMD" -c "
 import sys
@@ -412,12 +455,26 @@ with EvolutionCSV('$FULL_CSV_PATH') as csv:
   # Set current candidate for cleanup
   CURRENT_CANDIDATE_ID="$candidate_id"
   
-  # Process the candidate
-  if process_candidate "$candidate_id" "$parent_id" "$description"; then
+  # Process the candidate and capture exit code
+  process_candidate "$candidate_id" "$parent_id" "$description"
+  process_exit_code=$?
+  
+  if [[ $process_exit_code -eq 0 ]]; then
     echo "[WORKER-$$] Successfully processed $candidate_id"
+  elif [[ $process_exit_code -eq 77 ]]; then
+    # Special exit code 77 means AI failed to generate code
+    echo "[WORKER-$$] AI generation failed for $candidate_id - marking as failed-ai-retry"
+    # Mark with special status that indicates AI generation failed (not evaluation)
+    "$PYTHON_CMD" -c "
+import sys
+sys.path.insert(0, '$SCRIPT_DIR/..')
+from lib.evolution_csv import EvolutionCSV
+with EvolutionCSV('$FULL_CSV_PATH') as csv:
+    csv.update_candidate_status('$candidate_id', 'failed-ai-retry')
+" 2>/dev/null || true
   else
     echo "[WORKER-$$] Failed to process $candidate_id"
-    # Ensure status is set to failed (might already be done in process_candidate)
+    # Other failures (evaluation errors, etc) mark as failed
     "$PYTHON_CMD" -c "
 import sys
 sys.path.insert(0, '$SCRIPT_DIR/..')

package/lib/ai-cli.sh

@@ -19,6 +19,11 @@ call_ai_model_configured() {
       ai_output=$(timeout 300 claude --dangerously-skip-permissions --model "$model_name" -p "$prompt" 2>&1)
       local ai_exit_code=$?
       ;;
+    gpt-5)
+      local ai_output
+      ai_output=$(timeout 300 codex exec -m gpt-5 --dangerously-bypass-approvals-and-sandbox "$prompt" 2>&1)
+      local ai_exit_code=$?
+      ;;
     o3)
       local ai_output
       ai_output=$(timeout 300 codex exec -m o3 --dangerously-bypass-approvals-and-sandbox "$prompt" 2>&1)
@@ -100,7 +105,7 @@ clean_ai_output() {
   local model_name="$2"
   
   # Handle codex-specific output format
-  if [[ "$model_name" == "codex" || "$model_name" == "o3" ]]; then
+  if [[ "$model_name" == "codex" || "$model_name" == "o3" || "$model_name" == "gpt-5" ]]; then
     # Clean codex output - extract content between "codex" marker and "tokens used"
     if echo "$output" | grep -q "^\[.*\] codex$"; then
       # Extract content between "codex" line and "tokens used" line
@@ -191,7 +196,7 @@ call_ai_with_round_robin() {
     ai_output=$(call_ai_model_configured "$model" "$prompt")
     local ai_exit_code=$?
     
-    # Just check exit code - no interpretation
+    # Just check exit code
     if [[ $ai_exit_code -eq 0 ]]; then
       # Clean output if needed
       ai_output=$(clean_ai_output "$ai_output" "$model")

package/lib/config.sh

@@ -49,17 +49,13 @@ DEFAULT_AUTO_IDEATE=true
 # Default retry value
 DEFAULT_MAX_RETRIES=3
 
-# Default LLM CLI configuration (using eval for compatibility)
-declare -a DEFAULT_LLM_CLI_KEYS
-declare -a DEFAULT_LLM_CLI_VALUES
-DEFAULT_LLM_CLI_KEYS=(o3 codex gemini opus sonnet)
-DEFAULT_LLM_CLI_VALUES[0]='codex exec -m o3 --dangerously-bypass-approvals-and-sandbox "{{PROMPT}}"'
-DEFAULT_LLM_CLI_VALUES[1]='codex exec --dangerously-bypass-approvals-and-sandbox "{{PROMPT}}"'
-DEFAULT_LLM_CLI_VALUES[2]='gemini -y -p "{{PROMPT}}"'
-DEFAULT_LLM_CLI_VALUES[3]='claude --dangerously-skip-permissions --model opus -p "{{PROMPT}}"'
-DEFAULT_LLM_CLI_VALUES[4]='claude --dangerously-skip-permissions --model sonnet -p "{{PROMPT}}"'
-DEFAULT_LLM_RUN="sonnet"
-DEFAULT_LLM_IDEATE="gemini o3 opus"
+# Default memory limit (in MB, 0 means no limit)
+# Set to reasonable limit for ML workloads - about half of available system RAM
+DEFAULT_MEMORY_LIMIT_MB=12288
+
+# Default LLM CLI configuration - use simple variables instead of arrays
+DEFAULT_LLM_RUN="sonnet gpt-5 sonnet gpt-5"
+DEFAULT_LLM_IDEATE="gemini gpt-5 opus"
 
 # Load configuration from config file
 load_config() {
@@ -96,9 +92,13 @@ load_config() {
   # Set retry default
   MAX_RETRIES="$DEFAULT_MAX_RETRIES"
   
+  # Set memory limit default
+  MEMORY_LIMIT_MB="$DEFAULT_MEMORY_LIMIT_MB"
+  
   # Set LLM CLI defaults (compatibility for older bash)
   # Initialize associative array for LLM commands
   # Use simpler approach for compatibility
+  LLM_CLI_gpt_5='codex exec -m gpt-5 --dangerously-bypass-approvals-and-sandbox "{{PROMPT}}"'
   LLM_CLI_o3='codex exec -m o3 --dangerously-bypass-approvals-and-sandbox "{{PROMPT}}"'
   LLM_CLI_codex='codex exec --dangerously-bypass-approvals-and-sandbox "{{PROMPT}}"'
   LLM_CLI_gemini='gemini -y -p "{{PROMPT}}"'
@@ -202,12 +202,14 @@ load_config() {
           # Model definition - key is model name, value is command template
           # Remove single quotes from value if present
           value=$(echo "$value" | sed "s/^'//;s/'$//")
+          # Convert dashes to underscores for bash variable names
+          var_key=$(echo "$key" | sed 's/-/_/g')
           # Debug config loading
           if [[ "${DEBUG_CONFIG:-}" == "true" ]]; then
-            echo "[CONFIG DEBUG] Setting LLM_CLI_${key} = '$value'" >&2
+            echo "[CONFIG DEBUG] Setting LLM_CLI_${var_key} = '$value'" >&2
           fi
           # Use dynamic variable name for compatibility
-          eval "LLM_CLI_${key}=\"$value\""
+          eval "LLM_CLI_${var_key}=\"$value\""
         fi
       else
         # Handle top-level keys
@@ -221,6 +223,7 @@ load_config() {
           python_cmd) PYTHON_CMD="$value" ;;
           auto_ideate) AUTO_IDEATE="$value" ;;
           max_retries) MAX_RETRIES="$value" ;;
+          memory_limit_mb) MEMORY_LIMIT_MB="$value" ;;
           evolution_dir) 
             echo "[WARN] evolution_dir in config is ignored - automatically inferred from config file location" >&2
             ;;
@@ -316,14 +319,17 @@ show_config() {
   echo "  Lock timeout: $LOCK_TIMEOUT"
   echo "  Auto ideate: $AUTO_IDEATE"
   echo "  Max retries: $MAX_RETRIES"
+  echo "  Memory limit: ${MEMORY_LIMIT_MB}MB"
   echo "  LLM configuration:"
   # Show LLM configurations using dynamic variable names
-  for model in o3 codex gemini opus sonnet; do
+  for model in gpt_5 o3 codex gemini opus sonnet; do
     var_name="LLM_CLI_${model}"
     if [[ -n "${!var_name}" ]]; then
-      echo "    $model: ${!var_name}"
+      # Convert underscore back to dash for display
+      display_name=$(echo "$model" | sed 's/_/-/g')
+      echo "    $display_name: ${!var_name}"
     fi
   done
   echo "  LLM for run: $LLM_RUN"
   echo "  LLM for ideate: $LLM_IDEATE"
-}
+}

package/lib/csv_fixer.py

@@ -0,0 +1,35 @@
+#!/usr/bin/env python3
+"""
+CSV format fixer for claude-evolve
+Ensures proper quoting of CSV fields, especially descriptions
+"""
+
+import csv
+import sys
+
+def fix_csv_format(input_file, output_file):
+    """
+    Read a CSV file and ensure all fields are properly quoted.
+    The csv module handles quoting automatically based on content.
+    """
+    with open(input_file, 'r') as infile:
+        reader = csv.reader(infile)
+        rows = list(reader)
+    
+    with open(output_file, 'w', newline='') as outfile:
+        writer = csv.writer(outfile, quoting=csv.QUOTE_NONNUMERIC)
+        
+        # Write all rows - csv.writer handles quoting automatically
+        for row in rows:
+            writer.writerow(row)
+
+if __name__ == "__main__":
+    if len(sys.argv) != 3:
+        print("Usage: csv_fixer.py <input_file> <output_file>", file=sys.stderr)
+        sys.exit(1)
+    
+    try:
+        fix_csv_format(sys.argv[1], sys.argv[2])
+    except Exception as e:
+        print(f"Error fixing CSV: {e}", file=sys.stderr)
+        sys.exit(1)

package/lib/memory_limit_wrapper.py

@@ -0,0 +1,192 @@
+#!/usr/bin/env python3
+"""
+Memory-limited execution wrapper for claude-evolve evaluations.
+
+This script runs a command with memory limits to prevent runaway algorithms
+from consuming all system memory and crashing the machine.
+"""
+import sys
+import os
+import subprocess
+import signal
+import time
+import resource
+from typing import Optional
+
+def set_memory_limit(limit_mb: int) -> None:
+    """Set memory limit in MB using resource module."""
+    try:
+        # Convert MB to bytes
+        limit_bytes = limit_mb * 1024 * 1024
+        
+        # Set virtual memory limit (address space)
+        # On macOS this is the most reliable way to limit memory
+        resource.setrlimit(resource.RLIMIT_AS, (limit_bytes, limit_bytes))
+        
+        # Also try to set data segment limit if available
+        try:
+            resource.setrlimit(resource.RLIMIT_DATA, (limit_bytes, limit_bytes))
+        except (OSError, ValueError):
+            # Not available on all systems
+            pass
+            
+        print(f"[MEMORY] Set memory limit to {limit_mb}MB", file=sys.stderr)
+        
+    except (OSError, ValueError) as e:
+        print(f"[MEMORY] Warning: Could not set memory limit: {e}", file=sys.stderr)
+
+def monitor_memory_usage_native(process: subprocess.Popen, limit_mb: int) -> Optional[str]:
+    """Monitor process memory usage using native tools and kill if it exceeds limits."""
+    # print(f"[MEMORY] Starting native monitoring for PID {process.pid} with limit {limit_mb}MB", file=sys.stderr)
+    
+    while process.poll() is None:
+        try:
+            # Use ps command to get memory usage
+            ps_result = subprocess.run(
+                ["ps", "-o", "rss=", "-p", str(process.pid)], 
+                capture_output=True, 
+                text=True, 
+                timeout=1
+            )
+            
+            if ps_result.returncode == 0 and ps_result.stdout.strip():
+                # ps returns RSS in KB, convert to MB
+                memory_kb = int(ps_result.stdout.strip())
+                memory_mb = memory_kb / 1024
+                
+                # print(f"[MEMORY] PID {process.pid} using {memory_mb:.1f}MB (limit: {limit_mb}MB)", file=sys.stderr)
+                
+                if memory_mb > limit_mb:
+                    print(f"[MEMORY] Process exceeded {limit_mb}MB limit (using {memory_mb:.1f}MB), terminating", file=sys.stderr)
+                    # Kill the entire process group
+                    try:
+                        os.killpg(os.getpgid(process.pid), signal.SIGTERM)
+                        time.sleep(2)  # Give it time to cleanup
+                        if process.poll() is None:
+                            os.killpg(os.getpgid(process.pid), signal.SIGKILL)
+                    except ProcessLookupError:
+                        pass
+                    return f"Memory limit exceeded: {memory_mb:.1f}MB > {limit_mb}MB"
+            
+            time.sleep(0.5)  # Check every 500ms
+            
+        except (subprocess.TimeoutExpired, ValueError, ProcessLookupError):
+            # Process might have terminated or ps command failed
+            time.sleep(0.5)
+            continue
+    
+    # print(f"[MEMORY] Monitoring stopped for PID {process.pid}", file=sys.stderr)
+    return None
+
+def monitor_memory_usage(process: subprocess.Popen, limit_mb: int) -> Optional[str]:
+    """Monitor process memory usage and kill if it exceeds limits."""
+    try:
+        import psutil
+        ps_process = psutil.Process(process.pid)
+        
+        while process.poll() is None:
+            try:
+                # Get memory usage in MB
+                memory_info = ps_process.memory_info()
+                memory_mb = memory_info.rss / (1024 * 1024)
+                
+                if memory_mb > limit_mb:
+                    print(f"[MEMORY] Process exceeded {limit_mb}MB limit (using {memory_mb:.1f}MB), terminating", file=sys.stderr)
+                    # Kill the entire process group
+                    try:
+                        os.killpg(os.getpgid(process.pid), signal.SIGTERM)
+                        time.sleep(2)  # Give it time to cleanup
+                        if process.poll() is None:
+                            os.killpg(os.getpgid(process.pid), signal.SIGKILL)
+                    except ProcessLookupError:
+                        pass
+                    return f"Memory limit exceeded: {memory_mb:.1f}MB > {limit_mb}MB"
+                
+                time.sleep(0.5)  # Check every 500ms
+            except (psutil.NoSuchProcess, psutil.AccessDenied):
+                # Process already terminated
+                break
+    except ImportError:
+        # psutil not available, use native monitoring
+        return monitor_memory_usage_native(process, limit_mb)
+    
+    return None
+
+def main():
+    if len(sys.argv) < 3:
+        print("Usage: memory_limit_wrapper.py <memory_limit_mb> <command> [args...]", file=sys.stderr)
+        sys.exit(1)
+    
+    try:
+        memory_limit_mb = int(sys.argv[1])
+    except ValueError:
+        print(f"Error: Invalid memory limit '{sys.argv[1]}' - must be integer MB", file=sys.stderr)
+        sys.exit(1)
+    
+    command = sys.argv[2:]
+    
+    if memory_limit_mb <= 0:
+        print("[MEMORY] No memory limit set (0 or negative value)", file=sys.stderr)
+        # Just exec the command directly without limits
+        os.execvp(command[0], command)
+    
+    # Set memory limits for this process (inherited by subprocess)
+    set_memory_limit(memory_limit_mb)
+    
+    try:
+        # Start process in new process group for easier cleanup
+        process = subprocess.Popen(
+            command,
+            preexec_fn=os.setsid,  # Create new process group
+            stdout=subprocess.PIPE,
+            stderr=subprocess.STDOUT,
+            universal_newlines=True,
+            bufsize=1  # Line buffered
+        )
+        
+        # Monitor memory usage in background
+        memory_error = None
+        import threading
+        
+        def memory_monitor():
+            nonlocal memory_error
+            memory_error = monitor_memory_usage(process, memory_limit_mb)
+        
+        monitor_thread = threading.Thread(target=memory_monitor, daemon=True)
+        monitor_thread.start()
+        
+        # Stream output in real-time
+        while True:
+            output = process.stdout.readline()
+            if output == '' and process.poll() is not None:
+                break
+            if output:
+                print(output.rstrip())
+                sys.stdout.flush()
+        
+        # Wait for completion
+        return_code = process.wait()
+        
+        # Check if we killed it due to memory
+        if memory_error:
+            print(f"[MEMORY] {memory_error}", file=sys.stderr)
+            sys.exit(137)  # 128 + SIGKILL
+        
+        sys.exit(return_code)
+        
+    except FileNotFoundError:
+        print(f"Error: Command not found: {command[0]}", file=sys.stderr)
+        sys.exit(127)
+    except KeyboardInterrupt:
+        print("[MEMORY] Interrupted by user", file=sys.stderr)
+        try:
+            os.killpg(os.getpgid(process.pid), signal.SIGTERM)
+        except:
+            pass
+        sys.exit(130)
+    except Exception as e:
+        print(f"[MEMORY] Unexpected error: {e}", file=sys.stderr)
+        sys.exit(1)
+
+if __name__ == "__main__":
+    main()

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "claude-evolve",
-  "version": "1.5.3",
+  "version": "1.5.4",
   "bin": {
     "claude-evolve": "./bin/claude-evolve",
     "claude-evolve-main": "./bin/claude-evolve-main",

package/templates/config.yaml

@@ -46,6 +46,11 @@ auto_ideate: true
 # Maximum number of retries for failed candidates before marking as permanently failed
 max_retries: 3
 
+# Memory protection configuration
+# Memory limit in MB for evaluation processes (0 = no limit)
+# This prevents runaway algorithms from consuming all system memory
+memory_limit_mb: 2048
+
 # Parallel execution configuration
 parallel:
   # Enable parallel execution of evolution candidates
@@ -59,16 +64,11 @@ parallel:
 
 # LLM/AI CLI configuration
 llm_cli:
-  # How to run each CLI for each LLM option
-  # {{PROMPT}} will be replaced with the actual prompt text
-  o3: 'codex exec -m o3 --dangerously-bypass-approvals-and-sandbox "{{PROMPT}}"'
-  codex: 'codex exec --dangerously-bypass-approvals-and-sandbox "{{PROMPT}}"'
-  gemini: 'gemini -y -p "{{PROMPT}}"'
-  opus: 'claude --dangerously-skip-permissions --model opus -p "{{PROMPT}}"'
-  sonnet: 'claude --dangerously-skip-permissions --model sonnet -p "{{PROMPT}}"'
-
   # What to run for each sub-command
   # Models are tried in order, with round-robin distribution across candidates
   # You can repeat models for weighted selection (e.g., "sonnet sonnet gemini" for 2:1 ratio)
-  run: sonnet
-  ideate: gemini o3 opus
+
+  # commented out because these change over time; if you want to fix them in a particular
+  # configuration, uncomment them and set them
+  #run: sonnet
+  #ideate: gemini gpt-5 opus