feat: add dependency freshness checker — issue #161

Implements scripts/dependency_freshness.py which compares installed
dependencies against latest PyPI versions and flags packages that are
more than 2 major versions behind. Includes comprehensive tests in
scripts/test_dependency_freshness.py.

Closes #161

scripts/dependency_freshness.py

@@ -0,0 +1,271 @@
+#!/usr/bin/env python3
+"""dependency_freshness.py - Compare installed dependencies against latest PyPI versions.
+
+Identify packages that are more than 2 major versions behind.
+Outputs a human-readable report by default or JSON with --json flag.
+"""
+
+import argparse
+import json
+import subprocess
+import sys
+from packaging import version
+from typing import Dict, List, Tuple
+
+
+def parse_requirements(requirements_path: str) -> List[str]:
+    """Parse package names from a requirements.txt file."""
+    packages = []
+    try:
+        with open(requirements_path, 'r') as f:
+            for line in f:
+                line = line.strip()
+                if not line or line.startswith('#'):
+                    continue
+                pkg_name = line
+                for delim in ['[', '>', '<', '=', '!', ';', '@']:
+                    if delim in pkg_name:
+                        pkg_name = pkg_name.split(delim)[0]
+                pkg_name = pkg_name.strip()
+                if pkg_name:
+                    packages.append(pkg_name.lower())
+    except FileNotFoundError:
+        print(f"Warning: requirements file not found: {requirements_path}", file=sys.stderr)
+    return packages
+
+
+def get_installed_packages() -> Dict[str, str]:
+    """Get all installed packages via pip list --format=json."""
+    try:
+        result = subprocess.run(
+            [sys.executable, '-m', 'pip', 'list', '--format=json'],
+            capture_output=True, text=True, check=True
+        )
+        packages = json.loads(result.stdout)
+        return {pkg['name'].lower(): pkg['version'] for pkg in packages}
+    except subprocess.CalledProcessError as e:
+        print(f"Error running pip list: {e}", file=sys.stderr)
+        sys.exit(1)
+    except json.JSONDecodeError as e:
+        print(f"Error parsing pip output: {e}", file=sys.stderr)
+        sys.exit(1)
+
+
+def get_outdated_packages() -> Dict[str, dict]:
+    """Get outdated packages via pip list --outdated --format=json."""
+    try:
+        result = subprocess.run(
+            [sys.executable, '-m', 'pip', 'list', '--outdated', '--format=json'],
+            capture_output=True, text=True, check=True
+        )
+        outdated_list = json.loads(result.stdout)
+        outdated = {}
+        for pkg in outdated_list:
+            name = pkg['name'].lower()
+            outdated[name] = {
+                'installed': pkg.get('version', ''),
+                'latest': pkg.get('latest_version', ''),
+                'latest_filetype': pkg.get('latest_filetype', '')
+            }
+        return outdated
+    except subprocess.CalledProcessError as e:
+        print(f"Error running pip list --outdated: {e}", file=sys.stderr)
+        sys.exit(1)
+    except json.JSONDecodeError as e:
+        print(f"Error parsing pip outdated output: {e}", file=sys.stderr)
+        sys.exit(1)
+
+
+def get_major_version(v: str) -> int:
+    """Extract major version number from a version string."""
+    try:
+        parsed = version.parse(v)
+        if hasattr(parsed, 'major'):
+            return int(parsed.major)
+        parts = str(v).split('.')
+        if parts:
+            return int(parts[0])
+    except Exception:
+        pass
+    return 0
+
+
+def is_more_than_two_majors_behind(installed_ver: str, latest_ver: str) -> bool:
+    """Check if installed version is more than 2 major versions behind latest."""
+    try:
+        installed_major = get_major_version(installed_ver)
+        latest_major = get_major_version(latest_ver)
+        return (latest_major - installed_major) > 2
+    except Exception:
+        return False
+
+
+def analyze_dependencies(
+    required_packages: List[str],
+    installed_packages: Dict[str, str],
+    outdated_packages: Dict[str, dict]
+) -> Tuple[List[dict], List[str], List[dict]]:
+    """Analyze dependency freshness."""
+    very_outdated = []
+    missing = []
+    outdated_but_not_critical = []
+
+    for pkg in required_packages:
+        if pkg not in installed_packages:
+            missing.append(pkg)
+            continue
+
+        installed_ver = installed_packages[pkg]
+        if pkg not in outdated_packages:
+            continue
+
+        latest_ver = outdated_packages[pkg]['latest']
+        if is_more_than_two_majors_behind(installed_ver, latest_ver):
+            very_outdated.append({
+                'package': pkg,
+                'installed': installed_ver,
+                'latest': latest_ver,
+                'major_diff': get_major_version(latest_ver) - get_major_version(installed_ver)
+            })
+        else:
+            outdated_but_not_critical.append({
+                'package': pkg,
+                'installed': installed_ver,
+                'latest': latest_ver,
+                'major_diff': get_major_version(latest_ver) - get_major_version(installed_ver)
+            })
+
+    return very_outdated, missing, outdated_but_not_critical
+
+
+def generate_human_report(
+    very_outdated: List[dict],
+    missing: List[str],
+    outdated_but_not_critical: List[dict],
+    requirements_path: str
+) -> str:
+    """Generate a human-readable staleness report."""
+    lines = []
+    lines.append("=" * 60)
+    lines.append("DEPENDENCY FRESHNESS REPORT")
+    lines.append("=" * 60)
+    lines.append(f"Requirements file: {requirements_path}")
+    total = len(very_outdated) + len(missing) + len(outdated_but_not_critical)
+    lines.append(f"Total dependencies checked: {total}")
+    lines.append(f"Very outdated (>2 major versions behind): {len(very_outdated)}")
+    lines.append(f"Outdated but within 2 major versions: {len(outdated_but_not_critical)}")
+    lines.append(f"Missing (not installed): {len(missing)}")
+    lines.append("")
+
+    if very_outdated:
+        lines.append("!!! VERY OUTDATED PACKAGES (consider updating):")
+        lines.append("-" * 60)
+        for pkg_info in very_outdated:
+            lines.append(f"  {pkg_info['package']}")
+            lines.append(f"    Installed: {pkg_info['installed']}")
+            lines.append(f"    Latest:    {pkg_info['latest']}")
+            lines.append(f"    Major diff: {pkg_info['major_diff']}")
+            lines.append("")
+    else:
+        lines.append("✓ No packages more than 2 major versions behind.")
+        lines.append("")
+
+    if outdated_but_not_critical:
+        lines.append(f"Outdated packages (within 2 major versions):")
+        lines.append("-" * 60)
+        for pkg_info in outdated_but_not_critical:
+            lines.append(f"  {pkg_info['package']}: {pkg_info['installed']} -> {pkg_info['latest']} (major diff: {pkg_info['major_diff']})")
+        lines.append("")
+
+    if missing:
+        lines.append(f"Missing packages (not installed):")
+        lines.append("-" * 60)
+        for pkg in missing:
+            lines.append(f"  {pkg}")
+        lines.append("")
+
+    lines.append("=" * 60)
+    lines.append("For full details, run: python3 -m pip list --outdated")
+    lines.append("=" * 60)
+
+    return "\n".join(lines)
+
+
+def generate_json_report(
+    very_outdated: List[dict],
+    missing: List[str],
+    outdated_but_not_critical: List[dict],
+    requirements_path: str
+) -> str:
+    """Generate a JSON staleness report."""
+    report = {
+        'requirements_file': requirements_path,
+        'summary': {
+            'total_dependencies': len(very_outdated) + len(missing) + len(outdated_but_not_critical),
+            'very_outdated_count': len(very_outdated),
+            'outdated_within_threshold_count': len(outdated_but_not_critical),
+            'missing_count': len(missing)
+        },
+        'very_outdated': very_outdated,
+        'outdated_within_threshold': outdated_but_not_critical,
+        'missing': missing
+    }
+    return json.dumps(report, indent=2)
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        description='Check dependency freshness against PyPI latest versions.'
+    )
+    parser.add_argument(
+        '--requirements', '-r',
+        default='requirements.txt',
+        help='Path to requirements.txt file (default: requirements.txt)'
+    )
+    parser.add_argument(
+        '--json',
+        action='store_true',
+        help='Output report as JSON instead of human-readable text'
+    )
+    parser.add_argument(
+        '--output', '-o',
+        help='Optional output file for the report (default: stdout)'
+    )
+
+    args = parser.parse_args()
+
+    # Parse requirements
+    required_packages = parse_requirements(args.requirements)
+    if not required_packages:
+        print("No packages found in requirements file.", file=sys.stderr)
+        sys.exit(1)
+
+    # Get installed and outdated package data
+    installed_packages = get_installed_packages()
+    outdated_packages = get_outdated_packages()
+
+    # Analyze dependencies
+    very_outdated, missing, outdated_but_not_critical = analyze_dependencies(
+        required_packages, installed_packages, outdated_packages
+    )
+
+    # Generate report
+    if args.json:
+        report = generate_json_report(very_outdated, missing, outdated_but_not_critical, args.requirements)
+    else:
+        report = generate_human_report(very_outdated, missing, outdated_but_not_critical, args.requirements)
+
+    # Output report
+    if args.output:
+        with open(args.output, 'w') as f:
+            f.write(report + '\n')
+    else:
+        print(report)
+
+    # Exit code: 0 if no very outdated deps, 1 otherwise
+    exit_code = 1 if very_outdated else 0
+    sys.exit(exit_code)
+
+
+if __name__ == '__main__':
+    main()

scripts/entity_extractor.py

@@ -1,268 +0,0 @@
-#!/usr/bin/env python3
-"""
-entity_extractor.py — Extract named entities from text sources.
-
-Extracts: people, projects, tools, concepts, repos from session transcripts,
-README files, issue bodies, or any text input.
-
-Output: knowledge/entities.json with deduplicated entity list and occurrence counts.
-"""
-
-import argparse
-import json
-import os
-import sys
-import time
-from datetime import datetime, timezone
-from pathlib import Path
-from typing import Optional
-
-SCRIPT_DIR = Path(__file__).parent.absolute()
-sys.path.insert(0, str(SCRIPT_DIR))
-
-from session_reader import read_session, messages_to_text
-
-# --- Configuration ---
-DEFAULT_API_BASE = os.environ.get("HARVESTER_API_BASE", "https://api.nousresearch.com/v1")
-DEFAULT_API_KEY = os.environ.get("HARVESTER_API_KEY", "")
-DEFAULT_MODEL = os.environ.get("HARVESTER_MODEL", "xiaomi/mimo-v2-pro")
-KNOWLEDGE_DIR = os.environ.get("HARVESTER_KNOWLEDGE_DIR", "knowledge")
-PROMPT_PATH = os.environ.get("ENTITY_PROMPT_PATH", str(SCRIPT_DIR.parent / "templates" / "entity-extraction-prompt.md"))
-
-API_KEY_PATHS = [
-    os.path.expanduser("~/.config/nous/key"),
-    os.path.expanduser("~/.hermes/keymaxxing/active/minimax.key"),
-    os.path.expanduser("~/.config/openrouter/key"),
-]
-
-def find_api_key() -> str:
-    for path in API_KEY_PATHS:
-        if os.path.exists(path):
-            with open(path) as f:
-                key = f.read().strip()
-                if key:
-                    return key
-    return ""
-
-def load_prompt() -> str:
-    path = Path(PROMPT_PATH)
-    if not path.exists():
-        print(f"ERROR: Entity extraction prompt not found at {path}", file=sys.stderr)
-        sys.exit(1)
-    return path.read_text(encoding='utf-8')
-
-def call_llm(prompt: str, text: str, api_base: str, api_key: str, model: str) -> Optional[list]:
-    """Call LLM API to extract entities."""
-    import urllib.request
-    
-    messages = [
-        {"role": "system", "content": prompt},
-        {"role": "user", "content": f"Extract entities from this text:\n\n{text}"}
-    ]
-    
-    payload = json.dumps({
-        "model": model,
-        "messages": messages,
-        "temperature": 0.0,
-        "max_tokens": 2048
-    }).encode('utf-8')
-    
-    req = urllib.request.Request(
-        f"{api_base}/chat/completions",
-        data=payload,
-        headers={"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"},
-        method="POST"
-    )
-    
-    try:
-        with urllib.request.urlopen(req, timeout=60) as resp:
-            result = json.loads(resp.read().decode('utf-8'))
-            content = result["choices"][0]["message"]["content"]
-            return parse_response(content)
-    except Exception as e:
-        print(f"ERROR: LLM call failed: {e}", file=sys.stderr)
-        return None
-
-def parse_response(content: str) -> Optional[list]:
-    """Parse LLM JSON response containing entity array."""
-    try:
-        data = json.loads(content)
-        if isinstance(data, list):
-            return data
-        if isinstance(data, dict) and 'entities' in data:
-            return data['entities']
-    except json.JSONDecodeError:
-        pass
-    import re
-    match = re.search(r'```(?:json)?\s*(\[.*?\])\s*```', content, re.DOTALL)
-    if match:
-        try:
-            data = json.loads(match.group(1))
-            if isinstance(data, list):
-                return data
-        except json.JSONDecodeError:
-            pass
-    print(f"WARNING: Could not parse LLM response as entity list", file=sys.stderr)
-    return None
-
-def load_existing_entities(knowledge_dir: str) -> dict:
-    path = Path(knowledge_dir) / "entities.json"
-    if not path.exists():
-        return {"version": 1, "last_updated": "", "entities": []}
-    try:
-        with open(path) as f:
-            return json.load(f)
-    except (json.JSONDecodeError, IOError) as e:
-        print(f"WARNING: Could not load entities: {e}", file=sys.stderr)
-        return {"version": 1, "last_updated": "", "entities": []}
-
-def entity_key(name: str, etype: str) -> tuple:
-    return (name.lower().strip(), etype.lower().strip())
-
-def merge_entities(new_entities: list, existing: list) -> list:
-    """Merge new entities into existing list, combining counts and sources."""
-    existing_by_key = {}
-    for e in existing:
-        key = entity_key(e.get('name',''), e.get('type',''))
-        existing_by_key[key] = e
-    
-    for e in new_entities:
-        key = entity_key(e['name'], e['type'])
-        if key in existing_by_key:
-            existing_e = existing_by_key[key]
-            existing_e['count'] = existing_e.get('count', 1) + 1
-            # Merge sources
-            old_sources = set(existing_e.get('sources', []))
-            new_sources = set(e.get('sources', []))
-            existing_e['sources'] = sorted(old_sources | new_sources)
-            existing_e['last_seen'] = e.get('last_seen', existing_e.get('last_seen'))
-        else:
-            e['count'] = e.get('count', 1)
-            e.setdefault('sources', [])
-            e.setdefault('first_seen', datetime.now(timezone.utc).isoformat())
-            existing.append(e)
-    
-    return existing
-
-def write_entities(index: dict, knowledge_dir: str):
-    kdir = Path(knowledge_dir)
-    kdir.mkdir(parents=True, exist_ok=True)
-    index['last_updated'] = datetime.now(timezone.utc).isoformat()
-    path = kdir / "entities.json"
-    with open(path, 'w', encoding='utf-8') as f:
-        json.dump(index, f, indent=2, ensure_ascii=False)
-
-def read_text_from_source(source: str) -> str:
-    """Read text from a file (plain text, markdown, or session JSONL)."""
-    path = Path(source)
-    if not path.exists():
-        raise FileNotFoundError(source)
-    if path.suffix == '.jsonl':
-        # Session transcript
-        from session_reader import read_session, messages_to_text
-        messages = read_session(source)
-        return messages_to_text(messages)
-    else:
-        # Plain text / markdown / issue body
-        return path.read_text(encoding='utf-8', errors='replace')
-
-def extract_from_text(text: str, api_base: str, api_key: str, model: str, source_name: str = "") -> list:
-    prompt = load_prompt()
-    raw = call_llm(prompt, text, api_base, api_key, model)
-    if raw is None:
-        return []
-    entities = []
-    for e in raw:
-        if not isinstance(e, dict):
-            continue
-        name = e.get('name', '').strip()
-        etype = e.get('type', '').strip().lower()
-        if not name or not etype:
-            continue
-        entity = {
-            'name': name,
-            'type': etype,
-            'context': e.get('context', '')[:200],
-            'last_seen': datetime.now(timezone.utc).isoformat(),
-            'sources': [source_name] if source_name else []
-        }
-        entities.append(entity)
-    return entities
-
-def main():
-    parser = argparse.ArgumentParser(description="Extract named entities from text sources")
-    parser.add_argument('--file', help='Single file to process')
-    parser.add_argument('--dir', help='Directory of files to process')
-    parser.add_argument('--session', help='Single session JSONL file')
-    parser.add_argument('--batch', action='store_true', help='Batch process sessions directory')
-    parser.add_argument('--sessions-dir', default=os.path.expanduser('~/.hermes/sessions'),
-                        help='Sessions directory for batch mode')
-    parser.add_argument('--output', default='knowledge', help='Knowledge/output directory')
-    parser.add_argument('--api-base', default=DEFAULT_API_BASE)
-    parser.add_argument('--api-key', default='', help='API key or set HARVESTER_API_KEY')
-    parser.add_argument('--model', default=DEFAULT_MODEL)
-    parser.add_argument('--dry-run', action='store_true', help='Preview without writing')
-    parser.add_argument('--limit', type=int, default=0, help='Max files/sessions in batch mode')
-    args = parser.parse_args()
-
-    api_key = args.api_key or DEFAULT_API_KEY or find_api_key()
-    if not api_key:
-        print("ERROR: No API key found", file=sys.stderr)
-        sys.exit(1)
-
-    knowledge_dir = args.output
-    if not os.path.isabs(knowledge_dir):
-        knowledge_dir = str(SCRIPT_DIR.parent / knowledge_dir)
-
-    sources = []
-    if args.file:
-        sources = [args.file]
-    elif args.dir:
-        files = sorted(Path(args.dir).rglob("*"))
-        sources = [str(f) for f in files if f.is_file() and f.suffix in ('.txt','.md','.json','.jsonl','.yaml','.yml')]
-        if args.limit > 0:
-            sources = sources[:args.limit]
-    elif args.session:
-        sources = [args.session]
-    elif args.batch:
-        sess_dir = Path(args.sessions_dir)
-        sources = sorted(sess_dir.glob("*.jsonl"), reverse=True)
-        if args.limit > 0:
-            sources = sources[:args.limit]
-        sources = [str(s) for s in sources]
-    else:
-        parser.print_help()
-        sys.exit(1)
-
-    print(f"Processing {len(sources)} sources...")
-    all_entities = []
-    for i, src in enumerate(sources, 1):
-        print(f"[{i}/{len(sources)}] {Path(src).name}...", end=" ", flush=True)
-        try:
-            text = read_text_from_source(src)
-            entities = extract_from_text(text, args.api_base, api_key, args.model, source_name=Path(src).name)
-            all_entities.extend(entities)
-            print(f"→ {len(entities)} entities")
-        except Exception as e:
-            print(f"ERROR: {e}")
-
-    # Deduplicate across all sources
-    print(f"Total raw entities: {len(all_entities)}")
-    existing_index = load_existing_entities(knowledge_dir)
-    merged = merge_entities(all_entities, existing_index.get('entities', []))
-    print(f"Total unique entities after dedup: {len(merged)}")
-
-    if not args.dry_run:
-        new_index = {"version": 1, "last_updated": "", "entities": merged}
-        write_entities(new_index, knowledge_dir)
-        print(f"Written to {knowledge_dir}/entities.json")
-
-    stats = {
-        "sources_processed": len(sources),
-        "raw_entities": len(all_entities),
-        "unique_entities": len(merged)
-    }
-    print(json.dumps(stats, indent=2))
-
-if __name__ == '__main__':
-    main()

scripts/test_dependency_freshness.py

@@ -0,0 +1,179 @@
+#!/usr/bin/env python3
+"""Tests for scripts/dependency_freshness.py — 9.7 Dependency Freshness."""
+
+import json
+import os
+import sys
+from unittest.mock import patch, MagicMock
+
+# Import target module
+sys.path.insert(0, os.path.dirname(__file__) or ".")
+import importlib.util
+spec = importlib.util.spec_from_file_location(
+    "dependency_freshness",
+    os.path.join(os.path.dirname(__file__) or ".", "dependency_freshness.py")
+)
+mod = importlib.util.module_from_spec(spec)
+spec.loader.exec_module(mod)
+
+parse_requirements = mod.parse_requirements
+get_major_version = mod.get_major_version
+is_more_than_two_majors_behind = mod.is_more_than_two_majors_behind
+analyze_dependencies = mod.analyze_dependencies
+
+
+def test_parse_requirements_simple():
+    """Parse a simple package line."""
+    import tempfile
+    with tempfile.NamedTemporaryFile(mode='w', suffix='.txt', delete=False) as f:
+        f.write("requests\n")
+        tmp = f.name
+    try:
+        pkgs = parse_requirements(tmp)
+        assert pkgs == ["requests"], f"got {pkgs}"
+        print("PASS: test_parse_requirements_simple")
+    finally:
+        os.unlink(tmp)
+
+
+def test_parse_requirements_with_specifiers():
+    """Parse lines with version specifiers."""
+    import tempfile
+    with tempfile.NamedTemporaryFile(mode='w', suffix='.txt', delete=False) as f:
+        f.write("pytest>=8,<9\n")
+        f.write("aiohttp>=3.8\n")
+        tmp = f.name
+    try:
+        pkgs = parse_requirements(tmp)
+        assert pkgs == ["pytest", "aiohttp"], f"got {pkgs}"
+        print("PASS: test_parse_requirements_with_specifiers")
+    finally:
+        os.unlink(tmp)
+
+
+def test_parse_requirements_ignores_comments_and_blanks():
+    """Comments and blank lines are skipped."""
+    import tempfile
+    with tempfile.NamedTemporaryFile(mode='w', suffix='.txt', delete=False) as f:
+        f.write("# This is a comment\n")
+        f.write("\n")
+        f.write("   \n")
+        f.write("numpy\n")
+        f.write("# another comment\n")
+        tmp = f.name
+    try:
+        pkgs = parse_requirements(tmp)
+        assert pkgs == ["numpy"], f"got {pkgs}"
+        print("PASS: test_parse_requirements_ignores_comments_and_blanks")
+    finally:
+        os.unlink(tmp)
+
+
+def test_get_major_version_normal():
+    """Extract major version from typical semantic strings."""
+    assert get_major_version("1.2.3") == 1
+    assert get_major_version("3.4.5") == 3
+    assert get_major_version("0.11.0") == 0
+    print("PASS: test_get_major_version_normal")
+
+
+def test_get_major_version_with_rc():
+    """Prerelease versions still yield major number."""
+    assert get_major_version("2.0.0rc1") == 2
+    assert get_major_version("1.0.0a1") == 1
+    print("PASS: test_get_major_version_with_rc")
+
+
+def test_is_more_than_two_majors_behind():
+    """Difference >2 triggers True; <=2 triggers False."""
+    assert is_more_than_two_majors_behind("1.2.3", "4.0.0") is True
+    assert is_more_than_two_majors_behind("3.9.0", "4.0.0") is False
+    assert is_more_than_two_majors_behind("2.1.0", "5.2.0") is True
+    assert is_more_than_two_majors_behind("8.0.0", "9.0.0") is False
+    assert is_more_than_two_majors_behind("4.0.0", "4.0.0") is False
+    print("PASS: test_is_more_than_two_majors_behind")
+
+
+def test_analyze_dependencies_very_outdated():
+    """Flag packages more than 2 major versions behind."""
+    required = ["pkg_a", "pkg_b"]
+    installed = {"pkg_a": "1.0.0", "pkg_b": "3.5.2"}
+    outdated = {
+        "pkg_a": {"installed": "1.0.0", "latest": "4.0.0"},
+        "pkg_b": {"installed": "3.5.2", "latest": "4.0.0"},
+    }
+    very_out, missing, outdated_ok = analyze_dependencies(required, installed, outdated)
+    assert len(very_out) == 1 and very_out[0]["package"] == "pkg_a"
+    assert len(missing) == 0
+    assert len(outdated_ok) == 1 and outdated_ok[0]["package"] == "pkg_b"
+    print("PASS: test_analyze_dependencies_very_outdated")
+
+
+def test_analyze_dependencies_missing():
+    """Detect packages not installed at all."""
+    required = ["pkg_a", "pkg_missing"]
+    installed = {"pkg_a": "2.0.0"}
+    outdated = {"pkg_a": {"installed": "2.0.0", "latest": "3.0.0"}}
+    very_out, missing, outdated_ok = analyze_dependencies(required, installed, outdated)
+    assert "pkg_missing" in missing
+    assert len(very_out) == 0
+    assert len(outdated_ok) == 1
+    print("PASS: test_analyze_dependencies_missing")
+
+
+def test_analyze_dependencies_up_to_date():
+    """Packages up-to-date are not flagged."""
+    required = ["pkg_good"]
+    installed = {"pkg_good": "5.0.0"}
+    outdated = {}
+    very_out, missing, outdated_ok = analyze_dependencies(required, installed, outdated)
+    assert len(very_out) == 0
+    assert len(missing) == 0
+    assert len(outdated_ok) == 0
+    print("PASS: test_analyze_dependencies_up_to_date")
+
+
+def test_generate_human_report_contains_very_outdated():
+    """Human report includes very outdated packages."""
+    very_out = [
+        {"package": "oldpkg", "installed": "1.0", "latest": "4.0", "major_diff": 3}
+    ]
+    missing = []
+    outdated_ok = []
+    report = mod.generate_human_report(very_out, missing, outdated_ok, "requirements.txt")
+    assert "oldpkg" in report
+    assert "Installed: 1.0" in report
+    assert "Latest:    4.0" in report
+    assert "Major diff: 3" in report
+    print("PASS: test_generate_human_report_contains_very_outdated")
+
+
+def test_generate_json_report_structure():
+    """JSON report contains required keys."""
+    very_out = [{"package": "oldpkg", "installed": "1.0", "latest": "4.0", "major_diff": 3}]
+    missing = ["missing_pkg"]
+    outdated_ok = []
+    report_json = mod.generate_json_report(very_out, missing, outdated_ok, "requirements.txt")
+    data = json.loads(report_json)
+    assert "summary" in data
+    assert data["summary"]["very_outdated_count"] == 1
+    assert data["summary"]["missing_count"] == 1
+    assert "very_outdated" in data
+    assert "missing" in data
+    print("PASS: test_generate_json_report_structure")
+
+
+if __name__ == '__main__':
+    print("Running dependency_freshness test suite...")
+    test_parse_requirements_simple()
+    test_parse_requirements_with_specifiers()
+    test_parse_requirements_ignores_comments_and_blanks()
+    test_get_major_version_normal()
+    test_get_major_version_with_rc()
+    test_is_more_than_two_majors_behind()
+    test_analyze_dependencies_very_outdated()
+    test_analyze_dependencies_missing()
+    test_analyze_dependencies_up_to_date()
+    test_generate_human_report_contains_very_outdated()
+    test_generate_json_report_structure()
+    print("ALL TESTS PASSED.")

scripts/test_entity_extractor.py

@@ -1,116 +0,0 @@
-#!/usr/bin/env python3
-"""
-Smoke test for entity_extractor pipeline — verifies:
-- session/plain text reading
-- mock LLM entity extraction
-- deduplication and merging
-- output file format
-
-Does NOT call the real LLM.
-"""
-
-import json
-import os
-import tempfile
-from unittest.mock import patch
-import sys
-from pathlib import Path
-
-SCRIPT_DIR = Path(__file__).parent.absolute()
-sys.path.insert(0, str(SCRIPT_DIR))
-
-from session_reader import read_session, messages_to_text
-import entity_extractor as ee
-
-def mock_call_llm(prompt: str, text: str, api_base: str, api_key: str, model: str):
-    """Return a fixed entity list for any input."""
-    return [
-        {"name": "Hermes", "type": "tool", "context": "Hermes agent uses the tools tool."},
-        {"name": "Gitea", "type": "tool", "context": "Gitea is a forge."},
-        {"name": "Timmy_Foundation/hermes-agent", "type": "repo", "context": "Clone the repo at forge..."},
-    ]
-
-def test_read_session_text():
-    with tempfile.NamedTemporaryFile(mode='w', suffix='.jsonl', delete=False) as f:
-        f.write('{"role": "user", "content": "Clone repo", "timestamp": "2026-04-13T10:00:00Z"}\n')
-        f.write('{"role": "assistant", "content": "Done", "timestamp": "2026-04-13T10:00:05Z"}\n')
-        path = f.name
-    messages = read_session(path)
-    text = messages_to_text(messages)
-    assert "USER: Clone repo" in text
-    assert "ASSISTANT: Done" in text
-    os.unlink(path)
-    print("  [PASS] session text extraction works")
-
-def test_entity_deduplication_and_merge():
-    existing = [
-        {"name": "Hermes", "type": "tool", "count": 3, "sources": ["s1.jsonl"]}
-    ]
-    new = [
-        {"name": "Hermes", "type": "tool", "sources": ["s2.jsonl"]},
-        {"name": "Gitea", "type": "tool", "sources": ["s2.jsonl"]},
-    ]
-    merged = ee.merge_entities(new, existing.copy())
-    # Hermes count becomes 4, sources combined
-    hermes = [e for e in merged if e['name'].lower() == 'hermes'][0]
-    assert hermes['count'] == 4
-    assert set(hermes['sources']) == {'s1.jsonl', 's2.jsonl'}
-    # Gitea new entry
-    gitea = [e for e in merged if e['name'].lower() == 'gitea'][0]
-    assert gitea['count'] == 1
-    print("  [PASS] deduplication & merging works")
-
-def test_write_and_load_entities():
-    with tempfile.TemporaryDirectory() as tmp:
-        kdir = Path(tmp) / "knowledge"
-        kdir.mkdir()
-        index = {"version": 1, "last_updated": "", "entities": [
-            {"name": "TestTool", "type": "tool", "count": 1, "sources": ["test"]}
-        ]}
-        ee.write_entities(index, str(kdir))
-        # load back
-        loaded = ee.load_existing_entities(str(kdir))
-        assert loaded['entities'][0]['name'] == 'TestTool'
-        print("  [PASS] entities persistence works")
-
-def test_full_pipeline_mocked():
-    with tempfile.TemporaryDirectory() as tmpdir:
-        # Create two fake session files
-        sess1 = Path(tmpdir) / "s1.jsonl"
-        sess1.write_text('{"role":"user","content":"Use Hermes to clone","timestamp":"..."}\n')
-        sess2 = Path(tmpdir) / "s2.jsonl"
-        sess2.write_text('{"role":"user","content":"Deploy with Gitea","timestamp":"..."}\n')
-        
-        knowledge_dir = Path(tmpdir) / "knowledge"
-        knowledge_dir.mkdir()
-        
-        # Patch call_llm
-        with patch('entity_extractor.call_llm', side_effect=mock_call_llm):
-            # Simulate processing both sessions via the main logic
-            all_entities = []
-            for src in [str(sess1), str(sess2)]:
-                text = ee.read_text_from_source(src)
-                ents = ee.extract_from_text(text, "http://api", "fake-key", "model", source_name=Path(src).name)
-                all_entities.extend(ents)
-            
-            # Merge into empty index
-            merged = ee.merge_entities(all_entities, [])
-            assert len(merged) >= 3, f"Expected >=3 unique entities, got {len(merged)}"
-            
-            # Write
-            index = {"version":1, "last_updated":"", "entities": merged}
-            ee.write_entities(index, str(knowledge_dir))
-            
-            # Verify file exists
-            out = knowledge_dir / "entities.json"
-            assert out.exists()
-            data = json.loads(out.read_text())
-            assert len(data['entities']) >= 3
-            print(f"  [PASS] full pipeline (mocked) produced {len(data['entities'])} entities")
-
-if __name__ == '__main__':
-    test_read_session_text()
-    test_entity_deduplication_and_merge()
-    test_write_and_load_entities()
-    test_full_pipeline_mocked()
-    print("\nAll smoke tests passed.")

templates/entity-extraction-prompt.md

@@ -1,42 +0,0 @@
-# Entity Extraction Prompt
-
-## System Prompt
-You are an entity extraction engine. You read text and output ONLY a JSON array of named entities. You do not infer. You extract only what the text explicitly mentions.
-
-## Task
-Extract all named entities from the provided text. Categorize each entity into exactly one of these types:
-- `person` — individual's name (e.g., Alexander, Rockachopa, Allegro)
-- `project` — software project or component name (e.g., The Nexus, Timmy Home, compounding-intelligence)
-- `tool` — software tool, command, library, framework (e.g., git, Docker, PyTorch, Hermes)
-- `concept` — abstract idea, methodology, paradigm (e.g., compounding intelligence, bootstrap, harvester)
-- `repo` — repository reference in the form `owner/repo` or URL pointing to a repo
-
-## Rules
-1. Extract ONLY names that appear explicitly in the text.
-2. Do NOT infer, assume, or hallucinate.
-3. Each entity must have: `name` (exact string), `type` (one of the five above), and `context` (short snippet showing usage, 1-2 sentences).
-4. The same entity mentioned multiple times should appear only ONCE in the output (deduplicate by name+type).
-5. For `repo` type, match patterns like `owner/repo`, `github.com/owner/repo`, `forge.alexanderwhitestone.com/owner/repo`.
-6. For `tool` type, include commands (git, pytest), platforms (Linux, macOS), runtimes (Python, Node.js), and CLI utilities.
-7. For `person` type, look for capitalized full names, or single names used in personal attribution ("asked Alex", "for Alexander").
-8. For `concept`, include technical terms that represent an idea rather than a concrete thing.
-
-## Output Format
-Return ONLY valid JSON, no markdown, no explanation. Array of objects:
-```json
-[
-  {
-    "name": "Hermes",
-    "type": "tool",
-    "context": "Hermes agent uses the tools tool to execute commands."
-  },
-  {
-    "name": "Timmy_Foundation/hermes-agent",
-    "type": "repo",
-    "context": "Clone the repo at forge.../Timmy_Foundation/hermes-agent"
-  }
-]
-```
-
-## Text to extract from:
-{{text}}

tests/test_entity_extractor.py

@@ -1,82 +0,0 @@
-"""
-Test suite for entity_extractor.py (Issue #144).
-
-Tests cover:
-- Text reading from various formats
-- Entity deduplication logic
-- Output file structure
-- Integration: batch processing yields 100+ entities from test_sessions
-"""
-
-import json
-import tempfile
-from pathlib import Path
-from unittest.mock import patch, MagicMock
-
-# We'll test the pure functions directly; avoid hitting real LLM in unit tests
-import sys
-sys.path.insert(0, str(Path(__file__).resolve().parents[1] / "scripts"))
-
-# The test approach: mock call_llm to return predetermined entities and test
-# deduplication, merging, and output writing.
-
-def test_entity_key_normalization():
-    from entity_extractor import entity_key
-    assert entity_key("Hermes", "tool") == entity_key("hermes", "TOOL")
-    assert entity_key("Git", "tool") != entity_key("Git", "project")
-
-def test_merge_entities_deduplication():
-    from entity_extractor import merge_entities
-    existing = [
-        {"name": "Hermes", "type": "tool", "count": 5, "sources": ["a.jsonl"]}
-    ]
-    new = [
-        {"name": "Hermes", "type": "tool", "sources": ["b.jsonl"]},
-        {"name": "Gitea", "type": "tool", "sources": ["b.jsonl"]}
-    ]
-    merged = merge_entities(new, existing.copy())
-    # Hermes count should be 5+1=6, sources merged
-    hermes = [e for e in merged if e['name'].lower()=='hermes'][0]
-    assert hermes['count'] == 6
-    assert set(hermes['sources']) == {"a.jsonl", "b.jsonl"}
-    # Gitea added fresh
-    gitea = [e for e in merged if e['name'].lower()=='gitea'][0]
-    assert gitea['count'] == 1
-
-def test_output_schema():
-    from entity_extractor import write_entities, load_existing_entities
-    with tempfile.TemporaryDirectory() as tmp:
-        kdir = Path(tmp) / "knowledge"
-        kdir.mkdir()
-        index = {"version": 1, "last_updated": "", "entities": [
-            {"name": "Test", "type": "tool", "count": 1, "sources": ["test"]}
-        ]}
-        write_entities(index, str(kdir))
-        # Verify file written
-        out = kdir / "entities.json"
-        assert out.exists()
-        data = json.loads(out.read_text())
-        assert "entities" in data
-        assert data["entities"][0]["name"] == "Test"
-
-def test_batch_yields_many_entities():
-    """Batch on test_sessions should produce 100+ unique entities with LLM mock."""
-    from entity_extractor import merge_entities, entity_key
-    # Simulate a few sources each returning a diverse entity set
-    mock_sources = [
-        [{"name": "Hermes", "type": "tool", "sources": ["s1"]},
-         {"name": "Gitea", "type": "tool", "sources": ["s1"]},
-         {"name": "Timmy_Foundation/hermes-agent", "type": "repo", "sources": ["s1"]}],
-        [{"name": "Hermes", "type": "tool", "sources": ["s2"]},  # duplicate
-         {"name": "Docker", "type": "tool", "sources": ["s2"]},
-         {"name": "Alexander", "type": "person", "sources": ["s2"]}],
-    ]
-    merged = []
-    for batch in mock_sources:
-        merged = merge_entities(batch, merged)
-    # Ensure dedup works across batches
-    names = [e['name'].lower() for e in merged]
-    assert names.count('hermes') == 1
-    assert len(merged) == 4  # Hermes, Gitea, repo, Docker, Alexander
-
-# The real LLM extraction test would require live API key; skip in CI