refactor: consolidate skill docs, add anti-drift tests, and apply audit fixes

DLO-13: Restructure vlm-library-workflow skill as safety contract layer.
- Rewrite SKILL.md (69 lines): safety contract, execution threshold semantics,
  six-step high-risk loop, decision rules, phase skeleton
- Delete redundant references (cli-reference, workflow, command-recipes, dev-guide)
- Add triage.md (failure mapping + preflight) and dev-map.md (module→test mapping)
- Add tests/test_docs_consistency.py: 78 parametrized tests verifying documented
  vlm commands exist in CLI registry
- Add CSV path mismatch test to test_plan_review.py (4th safety gate path)
- Delete vlm-expert.skill (Gemini package, 7 months stale) and README Gemini section

DLO-2 audit fixes: rate limiter injection, symmetric quarantine categories,
review-plan safety gates, parser improvements, planner validation.

CLI modularization: commands/ directory with one module per command group.

src/vlm/reports.py

@@ -565,3 +565,70 @@ def _format_duration(duration_seconds: float) -> str:
         parts.append(f"{seconds}s")
     
     return " ".join(parts)
+
+
+def completeness_from_analysis(analysis_data: dict) -> list[SeasonCompleteness]:
+    """Build SeasonCompleteness records from analysis JSON."""
+    result: list[SeasonCompleteness] = []
+    for row in analysis_data.get("completeness", []) or []:
+        result.append(
+            SeasonCompleteness(
+                series_title=row["series_title"],
+                season=row["season"],
+                episodes_found=row["episodes_found"],
+                episodes_missing=row["episodes_missing"],
+            )
+        )
+    return result
+
+
+def duplicate_groups_from_analysis(analysis_data: dict) -> list[DuplicateGroup]:
+    """Build DuplicateGroup records from analysis JSON."""
+    groups: list[DuplicateGroup] = []
+    for dup in analysis_data.get("duplicates", []) or []:
+        identity_data = dup["identity"]
+        if identity_data["type"] == "movie":
+            identity = MovieIdentity(
+                title=identity_data["title"],
+                year=identity_data.get("year"),
+                confidence=1.0,
+                needs_review=False,
+                original_filename="",
+            )
+        else:
+            identity = SeriesIdentity(
+                title=identity_data["title"],
+                season=identity_data.get("season"),
+                episodes=identity_data.get("episodes", []),
+                confidence=1.0,
+                needs_review=False,
+                original_filename="",
+            )
+
+        quality_by_path = {
+            str(item.get("path", "")): item for item in dup.get("quality_comparison", [])
+        }
+        files: list[VideoFile] = []
+        for file_path in dup.get("files", []) or []:
+            quality = quality_by_path.get(str(file_path), {})
+            files.append(
+                VideoFile(
+                    path=Path(file_path),
+                    filename=Path(file_path).name,
+                    size_bytes=int(quality.get("size_bytes", 0) or 0),
+                    modified_timestamp=datetime.now(timezone.utc),
+                    category="",
+                    resolution=quality.get("resolution"),
+                    codec=quality.get("codec"),
+                    duration_seconds=quality.get("duration_seconds"),
+                    bitrate_kbps=quality.get("bitrate_kbps"),
+                )
+            )
+        groups.append(
+            DuplicateGroup(
+                identity=identity,
+                files=files,
+                quality_comparison=dup.get("quality_comparison", []),
+            )
+        )
+    return groups