feat: add Test Coverage Checker — 6.6

Add automated script that identifies changed source files, checks for
corresponding test changes, and reports coverage gaps.

Acceptance — #124:
  - Identifies changed source files  (git diff --name-only HEAD)
  - Checks for corresponding test changes (source→test file mapping)
  - Reports: code without tests       (lists uncovered sources)
  - Output: coverage gap             (structured text/JSON)

Closes #124
Task: 6.6 — Test Coverage Checker

scripts/coverage_checker.py

@@ -0,0 +1,169 @@
+#!/usr/bin/env python3
+"""
+Test Coverage Checker — 6.6
+
+Identifies changed source files, checks for corresponding test changes,
+and reports code without test coverage.
+
+Usage:
+    python3 scripts/test_coverage_checker.py
+    python3 scripts/test_coverage_checker.py --format json
+    python3 scripts/test_coverage_checker.py --compare HEAD~1  # Compare against a specific ref
+
+Acceptance:
+  - Identifies changed source files   (git diff --name-only HEAD)
+  - Checks for corresponding test changes (matches source→test file mapping)
+  - Reports: code without tests        (lists coverage gaps)
+  - Output: coverage gap              (structured text/JSON)
+"""
+
+import argparse
+import json
+import subprocess
+import sys
+from pathlib import Path
+from typing import List, Tuple, Optional
+
+REPO_ROOT = Path(__file__).resolve().parent.parent
+
+
+def run_git_diff(ref: str = "HEAD") -> List[str]:
+    """Return list of changed file paths relative to given ref."""
+    result = subprocess.run(
+        ["git", "diff", "--name-only", ref],
+        capture_output=True, text=True, cwd=REPO_ROOT
+    )
+    if result.returncode != 0:
+        print(f"ERROR: git diff failed: {result.stderr}")
+        sys.exit(1)
+    return [p for p in result.stdout.splitlines() if p.strip()]
+
+
+def is_source_file(path: str) -> bool:
+    """True if path is a Python source file (not test)."""
+    return path.endswith(".py") and not path.startswith("tests/") and "/test" not in Path(path).name
+
+
+def is_test_file(path: str) -> bool:
+    """True if path is a test file."""
+    if not path.endswith(".py"):
+        return False
+    name = Path(path).name
+    # Test files: test_*.py or *_test.py or in tests/ directory
+    return (name.startswith("test_") or name.endswith("_test.py") or path.startswith("tests/"))
+
+
+def source_to_test_path(src_path: str) -> str:
+    """
+    Map a source file path to its expected test file path.
+    Convention: scripts/<name>.py -> tests/test_<name>.py
+                <module>.py -> tests/test_<module>.py
+    """
+    name = Path(src_path).name
+    stem = Path(name).stem  # without .py
+    # Common mapping: script name -> test_ prefix in tests/
+    test_name = f"test_{stem}.py"
+    return str(Path("tests") / test_name)
+
+
+def test_file_exists() -> bool:
+    """Check if the test file exists in the repo."""
+    return (REPO_ROOT / test_rel).exists()
+
+
+def analyze_coverage(changed_files: List[str]) -> dict:
+    """
+    For each changed source file, check if corresponding test file also changed.
+    Returns structured coverage gap report.
+    """
+    changed_sources = [f for f in changed_files if is_source_file(f)]
+    changed_tests = [f for f in changed_files if is_test_file(f)]
+
+    # Build set of test file paths that changed (relative paths)
+    changed_test_set = set(changed_tests)
+
+    # Build coverage gap
+    uncovered_sources = []
+    covered_sources = []
+    for src in changed_sources:
+        coverage_entry = {"file": src}
+        # Check: does the corresponding test file also appear in changed files?
+        test_rel = source_to_test_path(src)
+        if test_rel in changed_test_set:
+            coverage_entry["status"] = "covered"
+            coverage_entry["test_file"] = test_rel
+            covered_sources.append(coverage_entry)
+        else:
+            coverage_entry["status"] = "missing"
+            coverage_entry["suggested_test"] = test_rel
+            uncovered_sources.append(coverage_entry)
+
+    return {
+        "repo": REPO_ROOT.name,
+        "changed_sources": len(changed_sources),
+        "changed_tests": len(changed_tests),
+        "covered_sources": len(covered_sources),
+        "uncovered_sources": len(uncovered_sources),
+        "coverage_ratio": (
+            len(covered_sources) / len(changed_sources)
+            if changed_sources else 1.0
+        ),
+        "covered": covered_sources,
+        "uncovered": uncovered_sources,
+        "all_changed": changed_files,
+    }
+
+
+def main():
+    parser = argparse.ArgumentParser(description="Test Coverage Checker")
+    parser.add_argument("--format", choices=["text", "json"], default="text",
+                        help="Output format")
+    parser.add_argument("--compare", default="HEAD",
+                        help="Git ref to compare against (default: HEAD)")
+    args = parser.parse_args()
+
+    # Step 1: Identify changed files
+    print(f"Scanning changes vs {args.compare}...")
+    changed_files = run_git_diff(args.compare)
+    if not changed_files:
+        print("No changed files detected.")
+        sys.exit(0)
+
+    # Step 2: Analyze coverage
+    report = analyze_coverage(changed_files)
+
+    if args.format == "json":
+        print(json.dumps(report, indent=2))
+        sys.exit(0)
+
+    # Text output
+    print("=" * 60)
+    print("  TEST COVERAGE CHECKER")
+    print("=" * 60)
+    print(f"  Repository:  {report['repo']}")
+    print(f"  Changed files total: {len(changed_files)}")
+    print(f"  Source files changed: {report['changed_sources']}")
+    print(f"  Test files changed:   {report['changed_tests']}")
+    print()
+    print(f"  Coverage (sources with test changes): {report['coverage_ratio']:.0%}")
+    print(f"    Covered:   {report['covered_sources']} source file(s)")
+    print(f"    Uncovered: {report['uncovered_sources']} source file(s)")
+    print()
+
+    if report["uncovered"]:
+        print("  COVERAGE GAP — Source files without corresponding test changes:")
+        print("  " + "-" * 54)
+        for item in report["uncovered"]:
+            print(f"    {item['file']}")
+            print(f"      Suggested test: {item['suggested_test']}")
+        print()
+        print("  ACTION: Write or update tests for the files above.")
+        sys.exit(1)  # Non-zero exit to flag coverage gap
+    else:
+        print("  All changed source files have corresponding test coverage.")
+
+    print("=" * 60)
+
+
+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_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_coverage_checker.py

@@ -0,0 +1,116 @@
+#!/usr/bin/env python3
+"""Tests for coverage_checker — Issue #124 acceptance validation."""
+
+import subprocess
+import sys
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).parent.parent / "scripts"))
+
+from coverage_checker import (
+    is_source_file,
+    is_test_file,
+    source_to_test_path,
+    analyze_coverage,
+)
+
+
+class TestSourceFileDetection:
+    def test_script_in_scripts_dir(self):
+        assert is_source_file("scripts/freshness.py") is True
+
+    def test_module_in_root(self):
+        assert is_source_file("knowledge_staleness_check.py") is True
+
+    def test_excludes_test_files(self):
+        assert is_source_file("tests/test_freshness.py") is False
+
+    def test_excludes_non_py(self):
+        assert is_source_file("README.md") is False
+
+
+class TestTestFileDetection:
+    def test_test_prefix(self):
+        assert is_test_file("tests/test_freshness.py") is True
+
+    def test_test_suffix(self):
+        assert is_test_file("scripts/freshness_test.py") is True
+
+    def test_regular_py_is_not_test(self):
+        assert is_test_file("scripts/freshness.py") is False
+
+
+class TestSourceToTestMapping:
+    def test_scripts_mapping(self):
+        assert source_to_test_path("scripts/freshness.py") == "tests/test_freshness.py"
+
+    def test_root_module_mapping(self):
+        assert source_to_test_path("knowledge_staleness_check.py") == "tests/test_knowledge_staleness_check.py"
+
+
+class TestAnalyzeCoverage:
+    def test_no_changes(self):
+        report = analyze_coverage([])
+        assert report["changed_sources"] == 0
+        assert report["uncovered_sources"] == 0
+        assert report["coverage_ratio"] == 1.0
+
+    def test_all_covered(self):
+        changed = [
+            "scripts/freshness.py",
+            "tests/test_freshness.py",
+            "scripts/dedup.py",
+            "tests/test_dedup.py",
+        ]
+        report = analyze_coverage(changed)
+        assert report["uncovered_sources"] == 0
+        assert report["covered_sources"] == 2
+
+    def test_gap_detected(self):
+        changed = [
+            "scripts/new_feature.py",
+            "README.md",
+        ]
+        report = analyze_coverage(changed)
+        assert report["uncovered_sources"] == 1
+        assert report["uncovered"][0]["file"] == "scripts/new_feature.py"
+        assert report["uncovered"][0]["suggested_test"] == "tests/test_new_feature.py"
+
+    def test_mixed_coverage(self):
+        changed = [
+            "scripts/covered.py",
+            "tests/test_covered.py",
+            "scripts/uncovered.py",
+        ]
+        report = analyze_coverage(changed)
+        assert report["covered_sources"] == 1
+        assert report["uncovered_sources"] == 1
+
+
+def run_all():
+    t = TestSourceFileDetection()
+    t.test_script_in_scripts_dir()
+    t.test_module_in_root()
+    t.test_excludes_test_files()
+    t.test_excludes_non_py()
+
+    t2 = TestTestFileDetection()
+    t2.test_test_prefix()
+    t2.test_test_suffix()
+    t2.test_regular_py_is_not_test()
+
+    t3 = TestSourceToTestMapping()
+    t3.test_scripts_mapping()
+    t3.test_root_module_mapping()
+
+    t4 = TestAnalyzeCoverage()
+    t4.test_no_changes()
+    t4.test_all_covered()
+    t4.test_gap_detected()
+    t4.test_mixed_coverage()
+
+    print("All 11 tests passed!")
+
+
+if __name__ == "__main__":
+    run_all()

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