code2llm 0.5.110__tar.gz → 0.5.112__tar.gz

{code2llm-0.5.110 → code2llm-0.5.112}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: code2llm
-Version: 0.5.110
+Version: 0.5.112
 Summary: High-performance Python code flow analysis with optimized TOON format - CFG, DFG, call graphs, and intelligent code queries
 Home-page: https://github.com/wronai/stts
 Author: STTS Project
@@ -67,11 +67,11 @@ Dynamic: requires-python
 
 ## AI Cost Tracking
 
-![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.110-blue) ![Python](https://img.shields.io/badge/python-3.9+-blue) ![License](https://img.shields.io/badge/license-Apache--2.0-green)
-![AI Cost](https://img.shields.io/badge/AI%20Cost-$7.50-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-53.7h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fqwen%2Fqwen3--coder--next-lightgrey)
+![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.112-blue) ![Python](https://img.shields.io/badge/python-3.9+-blue) ![License](https://img.shields.io/badge/license-Apache--2.0-green)
+![AI Cost](https://img.shields.io/badge/AI%20Cost-$7.50-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-55.0h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fqwen%2Fqwen3--coder--next-lightgrey)
 
-- 🤖 **LLM usage:** $7.5000 (154 commits)
-- 👤 **Human dev:** ~$5373 (53.7h @ $100/h, 30min dedup)
+- 🤖 **LLM usage:** $7.5000 (162 commits)
+- 👤 **Human dev:** ~$5505 (55.0h @ $100/h, 30min dedup)
 
 Generated on 2026-04-18 using [openrouter/qwen/qwen3-coder-next](https://openrouter.ai/qwen/qwen3-coder-next)
 
@@ -112,43 +112,6 @@ code2llm ./ -f all
 code2llm ./ -f context
 ```
 
-### Generating Individual Files
-
-```bash
-# Health diagnostics only
-code2llm ./ -f toon
-
-# Structural map only
-code2llm ./ -f map
-
-# Refactoring queue only
-code2llm ./ -f evolution
-
-# LLM narrative only
-code2llm ./ -f context
-
-# Call graph (mermaid + YAML)
-code2llm ./ -f mermaid
-
-# Call graph YAML only (no visualization)
-code2llm ./ -f calls
-
-# Flow diagrams only
-code2llm ./ -f flow
-
-# Standard YAML export
-code2llm ./ -f yaml
-
-# JSON export
-code2llm ./ -f json
-
-# Combined formats
-code2llm ./ -f toon,map,evolution,context
-
-# All core formats
-code2llm ./ -f all
-```
-
 ### Performance Options
 ```bash
 # Fast analysis for large projects
@@ -300,7 +263,6 @@ cat context.md | xclip -sel clip  # Linux
 - `flow.mmd` - Detailed control flow with complexity colors
 - `calls.mmd` - Simple call graph
 - `compact_flow.mmd` - High-level module view
-- `calls.yaml` - Structured call graph data (YAML format)
 - `*.png` - Pre-rendered images
 
 **Example usage**:
@@ -311,12 +273,6 @@ xdg-open flow.png  # Linux
 
 # Edit in Mermaid Live Editor
 # Copy content of .mmd files to https://mermaid.live
-
-# Generate only call graph files (mermaid + YAML)
-code2llm ./ -f mermaid
-
-# Generate calls.yaml standalone (no visualization files)
-code2llm ./ -f calls
 ```
 
 ## 🔍 Common Analysis Patterns
@@ -446,9 +402,9 @@ code2llm ./ -f yaml --separate-orphans
 
 **Generated by**: `code2llm ./ -f all --readme`  
 **Analysis Date**: 2026-04-18  
-**Total Functions**: 1014  
+**Total Functions**: 1041  
 **Total Classes**: 111  
-**Modules**: 131  
+**Modules**: 135  
 
 For more information about code2llm, visit: https://github.com/tom-sapletta/code2llm
 

{code2llm-0.5.110 → code2llm-0.5.112}/README.md

@@ -3,11 +3,11 @@
 
 ## AI Cost Tracking
 
-![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.110-blue) ![Python](https://img.shields.io/badge/python-3.9+-blue) ![License](https://img.shields.io/badge/license-Apache--2.0-green)
-![AI Cost](https://img.shields.io/badge/AI%20Cost-$7.50-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-53.7h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fqwen%2Fqwen3--coder--next-lightgrey)
+![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.112-blue) ![Python](https://img.shields.io/badge/python-3.9+-blue) ![License](https://img.shields.io/badge/license-Apache--2.0-green)
+![AI Cost](https://img.shields.io/badge/AI%20Cost-$7.50-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-55.0h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fqwen%2Fqwen3--coder--next-lightgrey)
 
-- 🤖 **LLM usage:** $7.5000 (154 commits)
-- 👤 **Human dev:** ~$5373 (53.7h @ $100/h, 30min dedup)
+- 🤖 **LLM usage:** $7.5000 (162 commits)
+- 👤 **Human dev:** ~$5505 (55.0h @ $100/h, 30min dedup)
 
 Generated on 2026-04-18 using [openrouter/qwen/qwen3-coder-next](https://openrouter.ai/qwen/qwen3-coder-next)
 
@@ -48,43 +48,6 @@ code2llm ./ -f all
 code2llm ./ -f context
 ```
 
-### Generating Individual Files
-
-```bash
-# Health diagnostics only
-code2llm ./ -f toon
-
-# Structural map only
-code2llm ./ -f map
-
-# Refactoring queue only
-code2llm ./ -f evolution
-
-# LLM narrative only
-code2llm ./ -f context
-
-# Call graph (mermaid + YAML)
-code2llm ./ -f mermaid
-
-# Call graph YAML only (no visualization)
-code2llm ./ -f calls
-
-# Flow diagrams only
-code2llm ./ -f flow
-
-# Standard YAML export
-code2llm ./ -f yaml
-
-# JSON export
-code2llm ./ -f json
-
-# Combined formats
-code2llm ./ -f toon,map,evolution,context
-
-# All core formats
-code2llm ./ -f all
-```
-
 ### Performance Options
 ```bash
 # Fast analysis for large projects
@@ -236,7 +199,6 @@ cat context.md | xclip -sel clip  # Linux
 - `flow.mmd` - Detailed control flow with complexity colors
 - `calls.mmd` - Simple call graph
 - `compact_flow.mmd` - High-level module view
-- `calls.yaml` - Structured call graph data (YAML format)
 - `*.png` - Pre-rendered images
 
 **Example usage**:
@@ -247,12 +209,6 @@ xdg-open flow.png  # Linux
 
 # Edit in Mermaid Live Editor
 # Copy content of .mmd files to https://mermaid.live
-
-# Generate only call graph files (mermaid + YAML)
-code2llm ./ -f mermaid
-
-# Generate calls.yaml standalone (no visualization files)
-code2llm ./ -f calls
 ```
 
 ## 🔍 Common Analysis Patterns
@@ -382,9 +338,9 @@ code2llm ./ -f yaml --separate-orphans
 
 **Generated by**: `code2llm ./ -f all --readme`  
 **Analysis Date**: 2026-04-18  
-**Total Functions**: 1014  
+**Total Functions**: 1041  
 **Total Classes**: 111  
-**Modules**: 131  
+**Modules**: 135  
 
 For more information about code2llm, visit: https://github.com/tom-sapletta/code2llm
 

{code2llm-0.5.110 → code2llm-0.5.112}/code2llm/__init__.py

@@ -8,7 +8,7 @@ Includes NLP Processing Pipeline for query normalization, intent matching,
 and entity resolution with multilingual support.
 """
 
-__version__ = "0.5.110"
+__version__ = "0.5.112"
 __author__ = "STTS Project"
 
 # Core analysis components (lightweight, always needed)

{code2llm-0.5.110 → code2llm-0.5.112}/code2llm/cli_exports/formats.py

@@ -71,8 +71,10 @@ def _export_project_yaml(args, result, output_dir: Path):
 
 def _export_project_toon(args, result, output_dir: Path):
     """Export project.toon.yaml directly from the current analysis result."""
+    from ..exporters.project_yaml.evolution import load_previous_evolution
+
     project_yaml_exporter = ProjectYAMLExporter()
-    prev_evolution = project_yaml_exporter._load_previous_evolution(output_dir / 'project.yaml')
+    prev_evolution = load_previous_evolution(output_dir / 'project.yaml')
     data = project_yaml_exporter._build_project_yaml(result, prev_evolution)
 
     exporter = ToonViewGenerator()
@@ -214,15 +216,15 @@ def _export_mermaid_pngs(args, output_dir: Path) -> None:
 
 
 def _export_calls(args, result, output_dir: Path):
-    """Export standalone calls.yaml (call graph as structured YAML).
+    """Export standalone calls.toon.yaml (call graph in toon format).
     
-    Generates calls.yaml without any Mermaid files — useful for programmatic
-    analysis of call graphs without visualization overhead.
+    Generates calls.toon.yaml in human-readable toon format with hubs, modules,
+    and edges sections — useful for programmatic analysis of call graphs.
     """
     yaml_exporter = YAMLExporter()
-    yaml_exporter.export_calls(result, str(output_dir / 'calls.yaml'))
+    yaml_exporter.export_calls_toon(result, str(output_dir / 'calls.toon.yaml'))
     if args.verbose:
-        print(f"  - CALLS (call graph YAML): {output_dir / 'calls.yaml'}")
+        print(f"  - CALLS (toon format): {output_dir / 'calls.toon.yaml'}")
 
 
 def _export_mermaid(args, result, output_dir: Path):
@@ -245,9 +247,9 @@ def _export_mermaid(args, result, output_dir: Path):
     exporter.export_call_graph(result, str(output_dir / 'calls.mmd'))
     exporter.export_compact(result, str(output_dir / 'compact_flow.mmd'))
 
-    # Export calls.yaml (structured call graph data)
+    # Export calls.toon.yaml (structured call graph data in toon format)
     yaml_exporter = YAMLExporter()
-    yaml_exporter.export_calls(result, str(output_dir / 'calls.yaml'))
+    yaml_exporter.export_calls_toon(result, str(output_dir / 'calls.toon.yaml'))
 
     if args.verbose:
         files = ['flow.mmd']
@@ -255,7 +257,7 @@ def _export_mermaid(args, result, output_dir: Path):
             files.append('flow_detailed.mmd')
         if getattr(args, 'flow_full', False):
             files.append('flow_full.mmd')
-        files.extend(['calls.mmd', 'compact_flow.mmd', 'calls.yaml'])
+        files.extend(['calls.mmd', 'compact_flow.mmd', 'calls.toon.yaml'])
         print(f"  - Mermaid: {output_dir}/*.mmd ({', '.join(files)})")
 
     _export_mermaid_pngs(args, output_dir)

code2llm-0.5.112/code2llm/exporters/project_yaml/__init__.py

@@ -0,0 +1,15 @@
+"""Project YAML Exporter — unified single source of truth for project diagnostics.
+
+This package splits the original monolithic exporter into focused modules:
+  - health: CC metrics, alerts, duplicates
+  - modules: per-file metrics, exports
+  - hotspots: high fan-out detection
+  - evolution: append-only history log
+
+Backward compatible: ProjectYAMLExporter can still be imported from here
+or from the original location.
+"""
+
+from .core import ProjectYAMLExporter
+
+__all__ = ["ProjectYAMLExporter"]

code2llm-0.5.112/code2llm/exporters/project_yaml/constants.py

@@ -0,0 +1,15 @@
+"""Constants and thresholds for project YAML export."""
+
+# Cyclomatic Complexity thresholds
+CC_CRITICAL = 10
+CC_WARNING = 15
+CC_ERROR = 20
+CC_SEVERE = 25
+
+# Fan-out thresholds
+FAN_OUT_THRESHOLD = 10
+FAN_OUT_ERROR = 15
+FAN_OUT_SEVERE = 20
+
+# Module size threshold
+GOD_MODULE_LINES = 500

code2llm-0.5.112/code2llm/exporters/project_yaml/core.py

@@ -0,0 +1,118 @@
+"""Core ProjectYAMLExporter class — orchestrates all builders."""
+
+from collections import defaultdict
+from datetime import datetime
+from pathlib import Path
+from typing import Any, Dict, List
+
+import yaml
+
+from code2llm.core.models import AnalysisResult
+from code2llm.core.config import LANGUAGE_EXTENSIONS
+from code2llm.exporters.base import Exporter
+
+from code2llm.exporters.toon.helpers import _is_excluded, _scan_line_counts, _rel_path
+from .constants import GOD_MODULE_LINES
+from .health import build_health
+from .modules import build_modules
+from .hotspots import build_hotspots, build_refactoring
+from .evolution import build_evolution, load_previous_evolution
+
+
+class ProjectYAMLExporter(Exporter):
+    """Export unified project.yaml — single source of truth for diagnostics.
+
+    Combines data from analysis.toon, project.toon.yaml, context.md, and evolution.toon.yaml
+    into one machine-parseable YAML file.
+    """
+
+    def export(self, result: AnalysisResult, output_path: str, **kwargs) -> None:
+        """Generate project.yaml from AnalysisResult.
+
+        If the file already exists, the evolution section is appended (not replaced).
+        """
+        output = Path(output_path)
+        output.parent.mkdir(parents=True, exist_ok=True)
+
+        # Load previous evolution entries if file exists
+        prev_evolution = load_previous_evolution(output)
+
+        data = self._build_project_yaml(result, prev_evolution)
+
+        with open(output, "w", encoding="utf-8") as f:
+            yaml.dump(data, f, default_flow_style=False, allow_unicode=True, sort_keys=False)
+
+    def _build_project_yaml(
+        self, result: AnalysisResult, prev_evolution: List[Dict]
+    ) -> Dict[str, Any]:
+        """Build complete project.yaml structure."""
+        line_counts = _scan_line_counts(result.project_path)
+        # Filter out venv/site-packages/etc — only count lines of non-excluded files
+        filtered_lines = {
+            k: v for k, v in line_counts.items()
+            if not _is_excluded(k)
+        }
+        total_lines = sum(filtered_lines.values()) // 2  # keys stored twice (abs + rel)
+
+        modules = build_modules(result, line_counts)
+        health = build_health(result, modules)
+        hotspots = build_hotspots(result)
+        refactoring = build_refactoring(result, modules, hotspots)
+        evolution = build_evolution(health, total_lines, prev_evolution)
+
+        return {
+            "version": "1",
+            "project": {
+                "name": Path(result.project_path).name if result.project_path else "unknown",
+                "repo": result.project_path or "",
+                "language": self._detect_primary_language(result),
+                "analyzed_at": datetime.now().strftime("%Y-%m-%dT%H:%M:%SZ"),
+                "tool": "code2llm",
+                "stats": {
+                    "files": len(set(
+                        fi.file for fi in result.functions.values()
+                        if not _is_excluded(fi.file)
+                    )) or len(result.modules),
+                    "lines": total_lines,
+                    "functions": len([
+                        f for f in result.functions.values()
+                        if not _is_excluded(f.file)
+                    ]),
+                    "classes": len([
+                        c for c in result.classes.values()
+                        if not _is_excluded(c.file)
+                    ]),
+                },
+            },
+            "health": health,
+            "modules": modules,
+            "hotspots": hotspots,
+            "refactoring": refactoring,
+            "evolution": evolution,
+        }
+
+    def _detect_primary_language(self, result: AnalysisResult) -> str:
+        """Detect the primary language of the project based on file counts."""
+        lang_counts: Dict[str, int] = defaultdict(int)
+
+        # Count files by language
+        for mi in result.modules.values():
+            if _is_excluded(mi.file):
+                continue
+            detected = False
+            for lang, extensions in LANGUAGE_EXTENSIONS.items():
+                if any(mi.file.endswith(ext) for ext in extensions):
+                    lang_counts[lang] += 1
+                    detected = True
+                    break
+            if not detected:
+                # Fallback to extension
+                ext = Path(mi.file).suffix.lstrip('.').lower()
+                if ext:
+                    lang_counts[ext] += 1
+
+        if not lang_counts:
+            return "unknown"
+
+        # Return the most common language
+        return max(lang_counts.items(), key=lambda x: x[1])[0]

code2llm-0.5.112/code2llm/exporters/project_yaml/evolution.py

@@ -0,0 +1,46 @@
+"""Evolution history builder for project.yaml."""
+
+from datetime import datetime
+from pathlib import Path
+from typing import Any, Dict, List
+
+import yaml
+
+
+def build_evolution(
+    health: Dict[str, Any],
+    total_lines: int,
+    prev_evolution: List[Dict],
+) -> List[Dict[str, Any]]:
+    """Build append-only evolution history."""
+    today = datetime.now().strftime("%Y-%m-%d")
+
+    new_entry = {
+        "date": today,
+        "cc_avg": health["cc_avg"],
+        "critical": health["critical_count"],
+        "lines": total_lines,
+        "note": "Automated analysis",
+    }
+
+    # Avoid duplicate entries for same date
+    evolution = [e for e in prev_evolution if e.get("date") != today]
+    evolution.append(new_entry)
+
+    return evolution
+
+
+def load_previous_evolution(output_path: Path) -> List[Dict]:
+    """Load previous evolution entries from existing project.yaml."""
+    if not output_path.exists():
+        return []
+    try:
+        with open(output_path, "r", encoding="utf-8") as f:
+            data = yaml.safe_load(f)
+        if isinstance(data, dict) and "evolution" in data:
+            evo = data["evolution"]
+            if isinstance(evo, list):
+                return evo
+    except Exception:
+        pass
+    return []

code2llm-0.5.112/code2llm/exporters/project_yaml/health.py

@@ -0,0 +1,103 @@
+"""Health metrics builder for project.yaml."""
+
+from collections import defaultdict
+from typing import Any, Dict, List
+
+from code2llm.core.models import AnalysisResult, ClassInfo, FunctionInfo
+from code2llm.exporters.toon.helpers import _is_excluded
+from .constants import CC_CRITICAL, CC_WARNING, CC_ERROR, CC_SEVERE, FAN_OUT_THRESHOLD, FAN_OUT_ERROR, FAN_OUT_SEVERE
+
+
+def build_health(result: AnalysisResult, modules: List[Dict]) -> Dict[str, Any]:
+    """Build health section with CC metrics, alerts, and issues."""
+    all_cc = []
+    for fi in result.functions.values():
+        if _is_excluded(fi.file):
+            continue
+        all_cc.append(fi.complexity.get("cyclomatic_complexity", 0))
+
+    cc_avg = round(sum(all_cc) / len(all_cc), 1) if all_cc else 0.0
+    critical_count = sum(1 for c in all_cc if c >= CC_CRITICAL)
+
+    # Detect cycles
+    proj_metrics = result.metrics.get("project", {})
+    cycles = proj_metrics.get("circular_dependencies", [])
+
+    # Detect duplicates
+    dup_count = count_duplicates(result)
+
+    # Build alerts
+    alerts = build_alerts(result)
+
+    return {
+        "cc_avg": cc_avg,
+        "critical_count": critical_count,
+        "critical_limit": CC_CRITICAL,
+        "duplicates": dup_count,
+        "cycles": len(cycles),
+        "alerts": alerts if alerts else [],
+    }
+
+
+def build_alerts(result: AnalysisResult) -> List[Dict[str, Any]]:
+    """Build list of health alerts for high CC and high fan-out."""
+    alerts = []
+    for qname, fi in result.functions.items():
+        if _is_excluded(fi.file):
+            continue
+        cc = fi.complexity.get("cyclomatic_complexity", 0)
+        if cc >= CC_WARNING:
+            display = fi.name
+            if fi.class_name:
+                display = f"{fi.class_name}.{fi.name}"
+            if cc >= CC_SEVERE:
+                severity = "critical"
+            elif cc >= CC_ERROR:
+                severity = "error"
+            else:
+                severity = "warning"
+            alerts.append({
+                "type": "cc_exceeded",
+                "target": display,
+                "value": cc,
+                "limit": CC_WARNING,
+                "severity": severity,
+            })
+
+    fan_alerts = []
+    for qname, fi in result.functions.items():
+        if _is_excluded(fi.file):
+            continue
+        fan_out = len(set(fi.calls))
+        if fan_out >= FAN_OUT_THRESHOLD:
+            display = fi.name
+            if fi.class_name:
+                display = f"{fi.class_name}.{fi.name}"
+            if fan_out >= FAN_OUT_SEVERE:
+                severity = "critical"
+            elif fan_out >= FAN_OUT_ERROR:
+                severity = "error"
+            else:
+                severity = "warning"
+            fan_alerts.append({
+                "type": "high_fan_out",
+                "target": display,
+                "value": fan_out,
+                "limit": FAN_OUT_THRESHOLD,
+                "severity": severity,
+            })
+
+    # Sort alerts by severity (critical first), then by value desc
+    sev_order = {"critical": 0, "error": 1, "warning": 2, "info": 3}
+    all_alerts = alerts + fan_alerts
+    all_alerts.sort(key=lambda a: (sev_order.get(a["severity"], 9), -a["value"]))
+    return all_alerts[:20]
+
+
+def count_duplicates(result: AnalysisResult) -> int:
+    """Count duplicate class names in different files."""
+    name_files: Dict[str, List[str]] = defaultdict(list)
+    for qname, ci in result.classes.items():
+        if not _is_excluded(ci.file):
+            name_files[ci.name].append(ci.file)
+    return sum(1 for files in name_files.values() if len(set(files)) > 1)

code2llm-0.5.112/code2llm/exporters/project_yaml/hotspots.py

@@ -0,0 +1,106 @@
+"""Hotspots and refactoring priorities builder for project.yaml."""
+
+from pathlib import Path
+from typing import Any, Dict, List
+
+from code2llm.core.models import AnalysisResult, FunctionInfo
+from code2llm.exporters.toon.helpers import _is_excluded
+from .constants import FAN_OUT_THRESHOLD, CC_WARNING, GOD_MODULE_LINES
+
+
+def build_hotspots(result: AnalysisResult) -> List[Dict[str, Any]]:
+    """Build hotspots list (high fan-out functions)."""
+    spots = []
+    for qname, fi in result.functions.items():
+        if _is_excluded(fi.file):
+            continue
+        fan_out = len(set(fi.calls))
+        if fan_out >= FAN_OUT_THRESHOLD:
+            display = fi.name
+            if fi.class_name:
+                display = f"{fi.class_name}.{fi.name}"
+            note = hotspot_note(fi, fan_out)
+            spots.append({
+                "name": display,
+                "fan_out": fan_out,
+                "note": note,
+            })
+    spots.sort(key=lambda s: s["fan_out"], reverse=True)
+    return spots[:10]
+
+
+def hotspot_note(fi: FunctionInfo, fan_out: int) -> str:
+    """Generate descriptive note for a hotspot."""
+    if "format" in fi.name.lower() or "dispatch" in fi.name.lower():
+        return f"{fan_out}-way dispatch"
+    if "export" in fi.name.lower():
+        return f"Export with {fan_out} outputs"
+    if "analyze" in fi.name.lower() or "process" in fi.name.lower():
+        return f"Analysis pipeline, {fan_out} stages"
+    if fi.docstring:
+        return fi.docstring[:80]
+    return f"Orchestrates {fan_out} calls"
+
+
+def build_refactoring(
+    result: AnalysisResult,
+    modules: List[Dict],
+    hotspots: List[Dict],
+) -> Dict[str, Any]:
+    """Build prioritized refactoring actions."""
+    priorities = []
+
+    # High CC functions → split
+    for qname, fi in result.functions.items():
+        if _is_excluded(fi.file):
+            continue
+        cc = fi.complexity.get("cyclomatic_complexity", 0)
+        if cc >= CC_WARNING:
+            display = fi.name
+            if fi.class_name:
+                display = f"{fi.class_name}.{fi.name}"
+            rel = _rel_path(fi.file, result.project_path)
+            priorities.append({
+                "action": f"Split {display} (CC={cc})",
+                "impact": "high" if cc >= 25 else "medium",
+                "effort": "low",
+                "module": Path(rel).name,
+            })
+
+    # Cycles → break
+    proj_metrics = result.metrics.get("project", {})
+    cycles = proj_metrics.get("circular_dependencies", [])
+    for cycle in cycles[:3]:
+        priorities.append({
+            "action": f"Break circular dependency: {' → '.join(str(c) for c in cycle) if isinstance(cycle, list) else str(cycle)}",
+            "impact": "medium",
+            "effort": "low",
+        })
+
+    # High fan-out → reduce
+    for spot in hotspots[:3]:
+        if spot["fan_out"] >= 15:
+            priorities.append({
+                "action": f"Reduce {spot['name']} fan-out (currently {spot['fan_out']})",
+                "impact": "medium",
+                "effort": "medium",
+            })
+
+    # God modules → split
+    for mod in modules:
+        if mod["lines"] >= GOD_MODULE_LINES:
+            priorities.append({
+                "action": f"Split god module {mod['path']} ({mod['lines']}L, {mod['classes']} classes)",
+                "impact": "high",
+                "effort": "high",
+            })
+
+    # Sort: high impact first, then low effort first
+    impact_order = {"high": 0, "medium": 1, "low": 2}
+    effort_order = {"low": 0, "medium": 1, "high": 2}
+    priorities.sort(key=lambda p: (
+        impact_order.get(p.get("impact", "low"), 9),
+        effort_order.get(p.get("effort", "medium"), 9),
+    ))
+
+    return {"priorities": priorities[:15]}