refactor CLI command modules and synchronize docs

AGENTS.md

@@ -2,7 +2,7 @@
 
 ## Project Structure & Module Organization
 - Core package lives in `src/vlm/`.
-- CLI entrypoint is `src/vlm/cli.py` (`vlm` console script). Commands use `pass_context` and `CLIContext` from `context.py`; some command logic lives in `commands/` (e.g. `scan`, `analyze`, `plan`).
+- CLI entrypoint is `src/vlm/cli.py` (`vlm` console script). Commands use `pass_context` and `CLIContext` from `context.py`; command logic is modularized in `commands/` (e.g. `scan`, `parse`, `enrich`, `analyze`, `plan`, `execute`).
 - Functional modules by concern: scanning (`scanner.py`), parsing (`parser.py`), enrichment (`enrichment.py`, `cache.py`, `providers/`), analysis/planning/execution (`analysis.py`, `planner.py`, `executor.py`), I/O helpers (`io.py`), utilities (`utils.py`), state/reporting/logging (`state.py`, `reports.py`, `logging_config.py`), config (`config.py`), models (`models.py`).
 - Tests live in `tests/` and mirror feature areas (e.g. `tests/test_scanner.py`, `tests/test_cli_state.py`, `tests/test_enrichment.py`).
 - Project metadata and tool config are in `pyproject.toml`.
@@ -36,3 +36,8 @@
 ## Security & Configuration Tips
 - Do not commit local paths, personal media metadata, or generated state/log artifacts.
 - Validate config changes against `vlm --help` and at least one end-to-end CLI flow before merging.
+
+
+## Documentation baseline
+- Updated to reflect refactor results as of 2026-02-16.
+- Canonical release notes are tracked in `CHANGELOG.md`.

ARCHITECTURE_REVIEW.md

@@ -1,3 +1,6 @@
+> [!NOTE]
+> Status: Historical snapshot. Current refactor results and validated baseline are tracked in `CHANGELOG.md` (updated 2026-02-16).
+
 # Video Library Manager - Architecture Review
 
 **Review Date**: 2026-02-13

AUDIT_FIX_PLAN.md

@@ -1,3 +1,6 @@
+> [!NOTE]
+> Status: Historical snapshot. Current refactor results and validated baseline are tracked in `CHANGELOG.md` (updated 2026-02-16).
+
 # Post-Audit Fixes Implementation Plan (2026-02-13)
 
 ## Objective

CHANGELOG.md

@@ -0,0 +1,38 @@
+# Changelog
+
+## 2026-02-16
+
+### Refactor Results
+
+- Modularized CLI command implementations:
+  - Added `src/vlm/commands/parse.py`
+  - Added `src/vlm/commands/enrich.py`
+  - Added `src/vlm/commands/execute.py`
+  - Refactored `src/vlm/cli.py` to delegate `parse`, `enrich`, `execute`, and `rollback` to command modules
+
+- Unified I/O layer for JSON handling:
+  - Added `load_json_file`, `save_json_file`, `save_analysis_json` in `src/vlm/io.py`
+  - Updated analyze/enrich paths to use unified I/O entry points
+
+- Parser/config consistency improvements:
+  - Added shared `DEFAULT_VIDEO_EXTENSIONS` in `src/vlm/config.py`
+  - Updated `src/vlm/parser.py` to use the shared config constant instead of local hardcoded defaults
+
+- Execution reliability hardening:
+  - Updated `src/vlm/executor.py` transaction logging to fail gracefully when `~/.vlm` is not writable (warn + continue)
+
+- Test stability improvements:
+  - Standardized test timestamps to timezone-aware UTC (`datetime.now(timezone.utc)`)
+  - Added missing `timezone` imports where required
+
+### Verification
+
+- Full test suite passed after refactor:
+  - `477 passed`
+
+### Documentation Sync
+
+- Updated repository documentation set to match the refactor baseline:
+  - Core docs: `README.md`, `AGENTS.md`, `CLAUDE.md`, `GEMINI.md`
+  - Skill docs: `skills/vlm-library-workflow/SKILL.md` and all references under `skills/vlm-library-workflow/references/`
+  - Legacy review/plan docs now explicitly marked as historical snapshots and redirected to `CHANGELOG.md` as the current source of truth

CLAUDE.md

@@ -1,5 +1,8 @@
 # CLAUDE.md
 
+## Documentation Status
+- Synchronized with refactor baseline on 2026-02-16 (see `CHANGELOG.md`).
+
 This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
 
 ## Project Overview
@@ -91,11 +94,11 @@ VLM follows a read-first, multi-stage pipeline:
 ### Module Organization
 - `cli.py` - Click-based CLI interface, global options, command registration
 - `context.py` - CLIContext (config, paths) and pass_context for commands
-- `commands/` - Command implementations (scan, analyze, plan)
+- `commands/` - Command implementations (scan, parse, enrich, analyze, plan, execute/rollback)
 - `scanner.py` - File discovery using system `find` command, metadata extraction via ffprobe
 - `parser.py` - Filename parsing using regex patterns (movies: title + year, series: SxxExx)
 - `enrichment.py` - Enrichment pipeline; `cache.py` - SQLite cache; `providers/` - TMDB etc.
-- `io.py` - Load/save identities JSON/CSV, plan/analysis input helpers
+- `io.py` - Unified JSON/CSV I/O helpers (including analysis writer and data adapters)
 - `utils.py` - UTC time, format_size, shared helpers
 - `analysis.py` - Completeness checking (episode gaps) and duplicate detection
 - `duplicate_resolve.py` - Duplicate group resolution (by_quality, by_reputation, first_seen, manual)

CODE_IMPROVEMENTS.md

@@ -1,3 +1,6 @@
+> [!NOTE]
+> Status: Historical snapshot. Current refactor results and validated baseline are tracked in `CHANGELOG.md` (updated 2026-02-16).
+
 # VLM 代码改进清单
 
 本文档记录对 Video Library Manager (VLM) 项目的代码审查发现的问题及对应解决方案。排除 AI/OpenAPI 相关问题。

FIX_PLAN.md

@@ -1,3 +1,6 @@
+> [!NOTE]
+> Status: Historical snapshot. Current refactor results and validated baseline are tracked in `CHANGELOG.md` (updated 2026-02-16).
+
 # Fix Plan (Verified Issues Only)
 
 ## Objective

GEMINI.md

@@ -1,5 +1,8 @@
 # GEMINI.md
 
+## Documentation Status
+- Synced with repository refactor baseline on 2026-02-16 (source of truth: `CHANGELOG.md`).
+
 This document provides a comprehensive overview of the Video Library Manager (VLM) project, intended to be used as instructional context for Gemini.
 
 ## Project Overview
@@ -48,19 +51,19 @@ The project uses `uv` for dependency management.
 
 The main entry point is the `vlm` command.
 
-*   Initialize configuration: `vlm config init`
-*   Scan the library: `vlm scan`
-*   Parse filenames: `vlm parse`
-*   Enrich metadata: `vlm enrich`
-*   Analyze the library: `vlm analyze`
-*   Generate a plan: `vlm plan`
-*   Execute the plan (dry-run): `vlm execute`
-*   Execute the plan (with confirmation): `vlm execute --confirm`
-*   Rollback the last execution: `vlm rollback`
+*   Initialize configuration: `uv run vlm config init`
+*   Scan the library: `uv run vlm scan`
+*   Parse filenames: `uv run vlm parse`
+*   Enrich metadata: `uv run vlm enrich`
+*   Analyze the library: `uv run vlm analyze`
+*   Generate a plan: `uv run vlm plan`
+*   Execute the plan (dry-run): `uv run vlm execute`
+*   Execute the plan (with confirmation): `uv run vlm execute --confirm`
+*   Rollback the last execution: `uv run vlm rollback`
 
 **Running tests:**
 
-*   Run all tests: `pytest`
+*   Run all tests: `uv run pytest`
 
 ## Development Conventions
 

IMPLEMENTATION_PLAN_2026-02-13.md

@@ -1,3 +1,6 @@
+> [!NOTE]
+> Status: Historical snapshot. Current refactor results and validated baseline are tracked in `CHANGELOG.md` (updated 2026-02-16).
+
 # VLM 改进实施计划（2026-02-13）
 
 ## 目标

IMPROVEMENT_RECOMMENDATIONS_2026-02-13.md

@@ -1,3 +1,6 @@
+> [!NOTE]
+> Status: Historical snapshot. Current refactor results and validated baseline are tracked in `CHANGELOG.md` (updated 2026-02-16).
+
 # VLM 项目与 Skill 改进建议（2026-02-13）
 
 ## 1. 评估范围与依据

README.md

@@ -1,5 +1,12 @@
 # Video Library Manager
 
+## Documentation Status
+
+- Last synchronized: **2026-02-16**
+- CLI command refactor landed (`parse`, `enrich`, `execute`, `rollback` logic moved to `src/vlm/commands/`).
+- Unified JSON I/O interfaces are available in `src/vlm/io.py`.
+- Full test baseline after refactor: **477 passed**.
+
 A Python-based CLI tool for managing personal video collections with a safety-first, human-in-the-loop approach.
 
 ## Features
@@ -806,8 +813,11 @@ src/vlm/
 ├── context.py          # CLIContext and pass_context for commands
 ├── commands/           # Command implementations
 │   ├── scan.py         # Scan command
+│   ├── parse.py        # Parse command
+│   ├── enrich.py       # Enrich command
 │   ├── analyze.py      # Analyze command
-│   └── plan.py         # Plan command
+│   ├── plan.py         # Plan command
+│   └── execute.py      # Execute/Rollback commands
 ├── scanner.py          # File discovery and metadata extraction
 ├── parser.py           # Filename parsing (titles, years, episodes)
 ├── enrichment.py       # Title/reputation enrichment pipeline
@@ -815,7 +825,7 @@ src/vlm/
 ├── providers/          # External metadata providers (TMDB, etc.)
 │   ├── base.py         # Provider interface
 │   └── tmdb.py         # TMDB API client
-├── io.py               # JSON/CSV load/save, load_analysis_json, plan/analysis input helpers
+├── io.py               # Unified JSON/CSV I/O helpers (load/save JSON, analysis writer, plan/analysis adapters)
 ├── utils.py            # UTC time, format_size, etc.
 ├── analysis.py         # Completeness and duplicate detection
 ├── duplicate_resolve.py # Duplicate group keep-index (by_quality, by_reputation, by_reputation_quality_time, first_seen, manual)

REVIEW_REPORT.md

@@ -1,3 +1,6 @@
+> [!NOTE]
+> Status: Historical snapshot. Current refactor results and validated baseline are tracked in `CHANGELOG.md` (updated 2026-02-16).
+
 # Code Review Report (Verified)
 
 ## Scope

TMDB_REFACTOR_PLAN.md

@@ -1,3 +1,6 @@
+> [!NOTE]
+> Status: Historical snapshot. Current refactor results and validated baseline are tracked in `CHANGELOG.md` (updated 2026-02-16).
+
 # TMDB Enrichment 重构执行计划
 
 ## 1. 目标与范围

VLM_PROJECT_AUDIT_REPORT.md

@@ -1,3 +1,6 @@
+> [!NOTE]
+> Status: Historical snapshot. Current refactor results and validated baseline are tracked in `CHANGELOG.md` (updated 2026-02-16).
+
 # Video Library Manager (VLM) 深度审计报告
 
 **报告版本**：1.0

codex_review.md

@@ -1,3 +1,6 @@
+> [!NOTE]
+> Status: Historical snapshot. Current refactor results and validated baseline are tracked in `CHANGELOG.md` (updated 2026-02-16).
+
 # Codex 架构复核报告
 
 评审日期: 2026-02-13  

skills/vlm-library-workflow/SKILL.md

@@ -1,3 +1,6 @@
+# Documentation Status
+- Synced with refactor baseline on 2026-02-16.
+
 ---
 name: vlm-library-workflow
 description: Operate and extend the Video Library Manager (`vlm`) with a safety-first, human-in-the-loop workflow across scan, parse, enrich, analyze, plan, review-plan, execute, rollback, and developer verification. Use when requests involve organizing a video library, producing or reviewing `inventory.csv`/`identities.json`/`analysis.json`/`plan.json`, tuning VLM config templates, resolving duplicates or episode gaps, running dry-run/confirm execution, recovering changes via rollback, explaining VLM CLI usage, or developing/modifying VLM features (parser, providers, planner, executor, commands, tests).

skills/vlm-library-workflow/references/cli-reference.md

@@ -1,3 +1,6 @@
+# Documentation Status
+- Updated to current CLI options on 2026-02-16.
+
 # VLM CLI Reference
 
 ## Configuration
@@ -11,7 +14,7 @@
 - `vlm enrich [--input JSON] [--output JSON] [--refresh-all]`: Fetch TMDB metadata.
 - `vlm analyze [--input JSON] [--output JSON]`: Find gaps and duplicates.
 - `vlm plan [--input JSON] [--analysis JSON] [--output JSON]`: Generate operations.
-- `vlm execute [--plan JSON] [--confirm]`: Move/Rename files.
+- `vlm execute [--plan JSON] [--confirm] [--yes] [--verbose-ops] [--safe-mode] [--preserve-directories]`: Move/Rename/Quarantine operations with safety guards.
 - `vlm rollback [--log PATH]`: Undo operations.
 
 ## Management & Reporting
@@ -22,5 +25,5 @@
 ## Important Config Options (`~/.vlm/config.yaml`)
 - `library_root`: Path to the video collection.
 - `templates`: Naming patterns for movies and series.
-- `plan.duplicate_keep`: Strategy for duplicates (`by_quality`, `by_reputation`, `first_seen`, `manual`).
+- `plan.duplicate_keep`: Strategy for duplicates (`by_quality`, `by_reputation`, `by_reputation_quality_time`, `first_seen`, `manual`).
 - `enrichment.api_keys`: TMDB and OpenAI keys.

skills/vlm-library-workflow/references/command-recipes.md

@@ -1,3 +1,6 @@
+# Documentation Status
+- Synced with refactor baseline on 2026-02-16.
+
 # VLM Command Recipes
 
 ## Baseline

skills/vlm-library-workflow/references/dev-guide.md

@@ -1,7 +1,10 @@
+# Documentation Status
+- Updated for the modular command architecture on 2026-02-16.
+
 # VLM Developer Guide
 
 ## Project Structure
-- `src/vlm/cli.py`: Entry point and command definitions.
+- `src/vlm/cli.py`: Entry point and thin command wrappers.
 - `src/vlm/parser.py`: Regex-based filename parsing logic.
 - `src/vlm/enrichment.py`: Pipeline for external metadata fetching.
 - `src/vlm/providers/`: API implementations (e.g., TMDB).
@@ -11,7 +14,7 @@
 ## Adding a New Command
 1. Create a new module in `src/vlm/commands/`.
 2. Define the command using `@click.command()`.
-3. Register it in `src/vlm/cli.py` using `main.add_command()`.
+3. Register it in `src/vlm/cli.py` with a Click-decorated function that delegates to the module implementation.
 
 ## Modifying the Parser
 - The parser uses a sequence of regex patterns in `src/vlm/parser.py`.
@@ -21,8 +24,8 @@
 ## Data Models
 See `src/vlm/models.py` for core data structures:
 - `VideoFile`: Basic file metadata.
-- `MediaIdentity`: Parsed and enriched information.
-- `PlanOperation`: Definition of a file move/rename/quarantine.
+- `MovieIdentity` / `SeriesIdentity`: Parsed/enriched identity records.
+- `FileOperation`: Definition of a move/rename/quarantine/no-op/preserve-directory operation.
 
 ## Testing
 - **Unit Tests**: `pytest`

skills/vlm-library-workflow/references/workflow.md

@@ -1,3 +1,6 @@
+# Documentation Status
+- Synced with refactor baseline on 2026-02-16.
+
 # VLM Workflow Guide
 
 This guide details the standard end-to-end process for organizing a video library using VLM.

src/vlm/cli.py

@@ -193,180 +193,20 @@ def parse(ctx: CLIContext, input: Path, output: Path, inventory: Optional[Path])
         vlm parse --input my_inventory.csv           # Custom input
         vlm parse --output parsed_identities.json    # Custom output
     """
-    from datetime import datetime, timezone
-    from vlm.parser import parse_movie, parse_series
-    from vlm.io import load_inventory_csv, save_identities_json
-
-    config = ctx.config
-    logger = ctx.logger
-
     try:
-        # Display parse start message
-        click.echo(f"Parsing identities from: {input}")
-
-        # Load video metadata from inventory if provided
-        path_to_metadata = {}
-        if inventory:
-            click.echo(f"Loading video metadata from: {inventory}")
-            inventory_files = load_inventory_csv(inventory)
-            path_to_metadata = {str(vf.path): vf for vf in inventory_files}
-            click.echo(f"Loaded metadata for {len(path_to_metadata)} files")
-
-        click.echo()
-
-        # Load inventory via unified I/O layer
-        inventory_files = load_inventory_csv(input)
-        video_files = [
-            {
-                'path': str(vf.path),
-                'filename': vf.filename,
-                'category': vf.category,
-            }
-            for vf in inventory_files
-        ]
-
-        click.echo(f"Loaded {len(video_files)} files from inventory")
-        click.echo()
-        
-        # Parse identities based on category
-        movie_identities = []
-        series_identities = []
-        anime_files = []
-        other_files = []
-
-        def get_video_metadata(file_path: str) -> dict:
-            """Extract video metadata from inventory if available."""
-            if not path_to_metadata:
-                return {}
-            vf = path_to_metadata.get(file_path)
-            if not vf:
-                return {}
-            return {
-                'size_bytes': vf.size_bytes,
-                'modified_timestamp': vf.modified_timestamp.isoformat(),
-                'resolution': vf.resolution,
-                'codec': vf.codec,
-                'duration_seconds': vf.duration_seconds,
-                'bitrate_kbps': vf.bitrate_kbps,
-            }
-
-        for vf in video_files:
-            filename = vf['filename']
-            category = vf['category']
-            file_path = vf['path']
-            video_metadata = get_video_metadata(file_path)
-
-            if category == 'movie':
-                identity = parse_movie(filename, extensions=config.video_extensions)
-                record = {
-                    'path': file_path,
-                    'filename': filename,
-                    'category': category,
-                    'title': identity.title,
-                    'year': identity.year,
-                    'confidence': identity.confidence,
-                    'needs_review': identity.needs_review
-                }
-                if video_metadata:
-                    record['video_metadata'] = video_metadata
-                movie_identities.append(record)
-
-            elif category == 'series':
-                identity = parse_series(filename, extensions=config.video_extensions)
-                record = {
-                    'path': file_path,
-                    'filename': filename,
-                    'category': category,
-                    'title': identity.title,
-                    'season': identity.season,
-                    'episodes': identity.episodes,
-                    'confidence': identity.confidence,
-                    'needs_review': identity.needs_review
-                }
-                if video_metadata:
-                    record['video_metadata'] = video_metadata
-                series_identities.append(record)
-            
-            elif category == 'anime':
-                # Anime files are not parsed in v1
-                anime_files.append({
-                    'path': vf['path'],
-                    'filename': filename,
-                    'category': category,
-                    'note': 'Anime parsing deferred in v1'
-                })
-            
-            else:
-                # Other files are not parsed
-                other_files.append({
-                    'path': vf['path'],
-                    'filename': filename,
-                    'category': category,
-                    'note': 'Not categorized for parsing'
-                })
-        
-        # Display parsing statistics
-        click.echo("Parsing complete!")
-        click.echo()
-        click.echo("Results by category:")
-        click.echo(f"  Movies: {len(movie_identities)}")
-        
-        # Count movies needing review
-        movies_need_review = sum(1 for m in movie_identities if m['needs_review'])
-        if movies_need_review > 0:
-            click.echo(f"    - Need review: {movies_need_review}")
-        
-        click.echo(f"  Series: {len(series_identities)}")
-        
-        # Count series needing review
-        series_need_review = sum(1 for s in series_identities if s['needs_review'])
-        if series_need_review > 0:
-            click.echo(f"    - Need review: {series_need_review}")
-        
-        click.echo(f"  Anime: {len(anime_files)} (not parsed in v1)")
-        click.echo(f"  Other: {len(other_files)} (not parsed)")
-        
-        # Save parsed identities to JSON
-        click.echo()
-        click.echo(f"Saving parsed identities to: {output}")
-        
-        # Ensure output directory exists
-        output.parent.mkdir(parents=True, exist_ok=True)
-        
-        # Build JSON structure
-        generation_timestamp = datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%S")
-
-        # Use v2 schema if video metadata was embedded
-        schema_version = "2.0" if path_to_metadata else "1.0"
-
-        identities_data = {
-            'vlm_schema_version': schema_version,
-            'metadata': {
-                'generated': generation_timestamp,
-                'source_inventory': str(input),
-                'total_files': len(video_files)
-            },
-            'movies': movie_identities,
-            'series': series_identities,
-            'anime': anime_files,
-            'other': other_files
-        }
-        
-        # Write JSON via unified I/O layer
-        save_identities_json(identities_data, output)
-        
-        click.echo(f"Parsed identities saved successfully!")
-        
-        logger.info(f"Parse completed: {len(movie_identities)} movies, {len(series_identities)} series, saved to {output}")
-        
+        from vlm.commands.parse import parse_cmd
+        parse_cmd(ctx, input, output, inventory)
     except FileNotFoundError:
         click.echo(f"Error: Input file not found: {input}", err=True)
-        logger.error(f"Input file not found: {input}")
+        ctx.logger.error(f"Input file not found: {input}")
         sys.exit(1)
-    
-    except Exception as e:
-        click.echo(f"Error during parsing: {e}", err=True)
-        logger.error(f"Parse failed: {e}", exc_info=True)
+    except ValueError as e:
+        click.echo(f"Error: {e}", err=True)
+        ctx.logger.error(f"Parse failed: {e}")
+        sys.exit(1)
+    except OSError as e:
+        click.echo(f"Error reading/writing files: {e}", err=True)
+        ctx.logger.error(f"Parse file I/O failed: {e}", exc_info=True)
         sys.exit(1)
 
 
@@ -424,150 +264,32 @@ def enrich(
     Applies incremental cache-backed enrichment to parsed identities and writes
     results back into identities JSON.
     """
-    import json
-    from vlm.enrichment import enrich_identities_data
-
-    config = ctx.config
-    logger = ctx.logger
-
-    if output is None:
-        output = input
-
     try:
-        click.echo(f"Enriching identities from: {input}")
-        click.echo(f"Output file: {output}")
-        click.echo()
-
-        with open(input, 'r', encoding='utf-8') as jsonfile:
-            identities_data = json.load(jsonfile)
-
-        if refresh_all and refresh_changed_only:
-            click.echo("Error: --refresh-all and --refresh-changed-only are mutually exclusive.", err=True)
-            sys.exit(1)
-
-        if timeout < 1:
-            click.echo("Error: --timeout must be >= 1", err=True)
-            sys.exit(1)
-
-        if retries < 0:
-            click.echo("Error: --retries must be >= 0", err=True)
-            sys.exit(1)
-
-        refresh_mode = "incremental"
-        if refresh_all:
-            refresh_mode = "refresh_all"
-        elif refresh_changed_only:
-            refresh_mode = "refresh_changed_only"
-
-        total = (
-            len(identities_data.get('movies', []))
-            + len(identities_data.get('series', []))
-            + len(identities_data.get('anime', []))
-        )
-
-        progress_state: dict[str, Optional[click.ProgressBar]] = {"bar": None}
-        progress_position = {"current": 0}
-        progress_bucket = {"value": -1}
-        is_tty = bool(getattr(sys.stderr, "isatty", lambda: False)())
-
-        def _enrich_progress(processed: int, total_count: int, metrics: dict[str, int]) -> None:
-            if total_count <= 0:
-                return
-
-            if is_tty and progress_state["bar"] is None:
-                bar = click.progressbar(
-                    length=total_count,
-                    label="Enriching records",
-                    show_pos=True,
-                )
-                progress_state["bar"] = bar.__enter__()
-
-            step = processed - progress_position["current"]
-            if step > 0 and progress_state["bar"] is not None:
-                progress_state["bar"].update(step)
-                progress_position["current"] = processed
-
-            if not is_tty:
-                percent = int(processed * 100 / total_count)
-                bucket = percent // 5
-                if bucket > progress_bucket["value"] or processed == total_count:
-                    progress_bucket["value"] = bucket
-                    click.echo(
-                        "Progress: "
-                        f"{processed}/{total_count} ({percent}%) "
-                        f"api_calls={metrics.get('api_calls', 0)} "
-                        f"cache_hits={metrics.get('cache_hits', 0)} "
-                        f"failed={metrics.get('failed', 0)}"
-                    )
-
-        try:
-            if total == 0:
-                click.echo("No movie/series/anime records found to enrich.")
-            enriched_data, stats = enrich_identities_data(
-                identities_data,
-                config,
-                refresh_mode=refresh_mode,
-                request_timeout=timeout,
-                retries=retries,
-                logger=logger,
-                progress_callback=_enrich_progress,
-            )
-        finally:
-            if progress_state["bar"] is not None:
-                progress_state["bar"].__exit__(None, None, None)
-                click.echo()
-
-        output.parent.mkdir(parents=True, exist_ok=True)
-        with open(output, 'w', encoding='utf-8') as jsonfile:
-            json.dump(enriched_data, jsonfile, indent=2, ensure_ascii=False)
-
-        total_records = int(stats['total']) if stats['total'] else 0
-        cache_hits = int(stats['cache_hits'])
-        hit_rate = (cache_hits / total_records * 100.0) if total_records else 0.0
-
-        click.echo("Enrichment complete!")
-        click.echo(f"  Total records: {total_records}")
-        click.echo(f"  Refresh mode: {refresh_mode}")
-        click.echo(f"  Enriched now: {stats['enriched']}")
-        click.echo(f"  Cache hits: {stats['cache_hits']}")
-        click.echo(f"  Cache hit rate: {hit_rate:.1f}%")
-        click.echo(f"  API calls: {stats['api_calls']}")
-        click.echo(f"  Failed requests: {stats['failed']}")
-        click.echo(f"  Skipped: {stats['skipped']}")
-        click.echo(f"  Needs review: {stats['needs_review']}")
-        skip_reasons = stats.get("skip_reasons", {})
-        if isinstance(skip_reasons, dict):
-            non_zero = [f"{name}={count}" for name, count in sorted(skip_reasons.items()) if int(count) > 0]
-            if non_zero:
-                click.echo(f"  Skip reasons: {' '.join(non_zero)}")
-        failed_items = stats.get('failed_items', [])
-        if isinstance(failed_items, list) and failed_items:
-            click.echo("  Failure sample:")
-            for item in failed_items[:3]:
-                click.echo(
-                    f"    - [{item.get('provider', 'unknown')}] {item.get('title', '')}: {item.get('reason', '')}"
-                )
-
-        logger.info(
-            "Enrich completed: total=%s enriched=%s cache_hits=%s failed=%s skipped=%s mode=%s",
-            stats['total'],
-            stats['enriched'],
-            stats['cache_hits'],
-            stats['failed'],
-            stats['skipped'],
-            refresh_mode,
+        from vlm.commands.enrich import enrich_cmd
+        enrich_cmd(
+            ctx,
+            input,
+            output,
+            refresh_changed_only,
+            refresh_all,
+            timeout,
+            retries,
         )
     except FileNotFoundError:
         click.echo(f"Error: Input file not found: {input}", err=True)
-        logger.error(f"Input file not found: {input}")
+        ctx.logger.error(f"Input file not found: {input}")
         sys.exit(1)
     except json.JSONDecodeError as e:
         click.echo(f"Error: Failed to parse JSON file: {e}", err=True)
-        logger.error(f"JSON parsing failed during enrich: {e}", exc_info=True)
+        ctx.logger.error(f"JSON parsing failed during enrich: {e}", exc_info=True)
         sys.exit(1)
-    except Exception as e:
-        click.echo(f"Error during enrichment: {e}", err=True)
-        logger.error(f"Enrich failed: {e}", exc_info=True)
+    except ValueError as e:
+        click.echo(f"Error: {e}", err=True)
+        ctx.logger.error(f"Enrich validation failed: {e}")
+        sys.exit(1)
+    except OSError as e:
+        click.echo(f"Error reading/writing files: {e}", err=True)
+        ctx.logger.error(f"Enrich file I/O failed: {e}", exc_info=True)
         sys.exit(1)
 
 
@@ -791,8 +513,20 @@ def review_plan_cmd(
     default=False,
     help='Print per-operation dry-run logs at INFO level'
 )
+@click.option(
+    '--preserve-directories',
+    is_flag=True,
+    default=False,
+    help='Preserve empty source directories instead of allowing them to be destroyed'
+)
+@click.option(
+    '--safe-mode',
+    is_flag=True,
+    default=False,
+    help='Enable safe mode: prevent any operations that would destroy directories'
+)
 @pass_context
-def execute(ctx: CLIContext, plan: Path, confirm: bool, yes: bool, verbose_ops: bool):
+def execute(ctx: CLIContext, plan: Path, confirm: bool, yes: bool, verbose_ops: bool, preserve_directories: bool, safe_mode: bool):
     """Execute plan (defaults to dry-run, requires --confirm).
 
     Executes file operations from a plan. Defaults to dry-run mode which
@@ -806,155 +540,20 @@ def execute(ctx: CLIContext, plan: Path, confirm: bool, yes: bool, verbose_ops:
         vlm execute --confirm            # Actually execute operations (with prompt)
         vlm execute --confirm --yes      # Execute without confirmation prompt
     """
-    from vlm.planner import load_plan
-    from vlm.executor import ExecutionEngine
-    
-    config = ctx.config
-    logger = ctx.logger
-    
     try:
-        # Determine execution mode
-        mode = "execute" if confirm else "dry-run"
-        
-        # Display execution start message
-        click.echo(f"Loading execution plan from: {plan}")
-        click.echo()
-        
-        # Load execution plan
-        execution_plan = load_plan(plan)
-        
-        # Display plan summary
-        click.echo(f"Execution plan loaded: {execution_plan.plan_id}")
-        click.echo(f"Created at: {execution_plan.created_at}")
-        click.echo(f"Total operations: {len(execution_plan.operations)}")
-        if execution_plan.human_summary:
-            click.echo()
-            click.echo(execution_plan.human_summary)
-        elif execution_plan.summary or execution_plan.summary_by_reason:
-            s = execution_plan.summary or {}
-            by_r = execution_plan.summary_by_reason or {}
-            parts = [f"操作统计：共 {s.get('total', len(execution_plan.operations))} 条（move {s.get('move', 0)}，rename {s.get('rename', 0)}，quarantine {s.get('quarantine', 0)}，no-op {s.get('no-op', 0)}）"]
-            if by_r:
-                parts.append("原因分布：" + "；".join(f"{r}: {c}" for r, c in list(by_r.items())[:5]))
-            click.echo()
-            click.echo("\n".join(parts))
-        click.echo()
-        
-        # Display mode warning
-        if mode == "dry-run":
-            click.echo("⚠️  DRY-RUN MODE - No files will be modified")
-            click.echo("   Use --confirm to actually execute operations")
-        else:
-            click.echo("⚠️  EXECUTE MODE - Files will be modified!")
-            click.echo("   This operation cannot be undone without rollback")
-            click.echo()
-            if not yes:
-                if not click.confirm("Are you sure you want to proceed?"):
-                    click.echo("Execution cancelled.")
-                    return
-            else:
-                click.echo("Auto-approved via --yes flag")
-        
-        click.echo()
-        click.echo(f"Executing {len(execution_plan.operations)} operations...")
-        click.echo()
-        
-        # Load state manager to update file statuses during execution
-        from vlm.state import StateManager
-        state_path = Path.home() / ".vlm" / "state.json"
-        state_manager = StateManager(state_path)
-
-        # Create execution engine and execute plan
-        engine = ExecutionEngine(
-            logger=logger, 
-            config=config, 
-            verbose_operations=verbose_ops,
-            state_manager=state_manager
-        )
-        results, summary, rollback_log = engine.execute_plan(
-            execution_plan,
-            mode=mode,
-            confirmed=confirm
-        )
-        
-        # Display execution progress (show some operations)
-        if mode == "dry-run":
-            click.echo("Sample operations (dry-run):")
-            # Show first 5 operations as examples
-            for i, result in enumerate(results[:5]):
-                op = result.operation
-                if op.operation_type != "no-op":
-                    click.echo(f"  [{i+1}] {op.operation_type}: {op.source_path.name}")
-                    if op.destination_path:
-                        click.echo(f"      -> {op.destination_path}")
-            
-            if len(results) > 5:
-                click.echo(f"  ... and {len(results) - 5} more operations")
-        else:
-            # In execute mode, show progress for all operations
-            for i, result in enumerate(results):
-                op = result.operation
-                if op.operation_type != "no-op" and not op.has_conflict:
-                    status = "✓" if result.success else "✗"
-                    click.echo(f"  [{i+1}/{len(results)}] {status} {op.operation_type}: {op.source_path.name}")
-                    if result.error_message:
-                        click.echo(f"      Error: {result.error_message}")
-        
-        # Display execution summary
-        click.echo()
-        click.echo("=" * 60)
-        click.echo(f"Execution Summary ({mode} mode)")
-        click.echo("=" * 60)
-        click.echo(f"  Total operations: {summary['total']}")
-        click.echo(f"  Successful: {summary['successful']}")
-        click.echo(f"  Failed: {summary['failed']}")
-        click.echo(f"  Skipped: {summary['skipped']}")
-        click.echo()
-        
-        # Save rollback log if in execute mode
-        if mode == "execute" and rollback_log:
-            # Save to ~/.vlm/rollback/ directory
-            rollback_dir = Path.home() / ".vlm" / "rollback"
-            rollback_dir.mkdir(parents=True, exist_ok=True)
-            rollback_path = rollback_dir / f"rollback_{rollback_log.log_id}.json"
-            
-            engine.save_rollback_log(rollback_log, rollback_path)
-            
-            click.echo(f"Rollback log saved to: {rollback_path}")
-            click.echo()
-            click.echo("To undo these operations, run:")
-            click.echo(f"  vlm rollback --log {rollback_path}")
-            click.echo()
-        
-        # Log completion
-        if mode == "dry-run":
-            click.echo("Dry-run complete! No files were modified.")
-            click.echo("Review the operations above and use --confirm to execute.")
-        else:
-            if summary['failed'] > 0:
-                click.echo(f"⚠️  Execution completed with {summary['failed']} failures.")
-                click.echo("   Check the log file for details.")
-            else:
-                click.echo("✓ Execution completed successfully!")
-        
-        logger.info(
-            f"Execution completed in {mode} mode: "
-            f"{summary['successful']} successful, {summary['failed']} failed, {summary['skipped']} skipped"
-        )
-        
+        from vlm.commands.execute import execute_cmd
+        execute_cmd(ctx, plan, confirm, yes, verbose_ops, preserve_directories, safe_mode)
     except FileNotFoundError:
-        click.echo(f"Error: Plan file not found: {plan}", err=True)
-        logger.error(f"Plan file not found: {plan}")
+        click.echo(f"Error: File not found: {plan}", err=True)
+        ctx.logger.error(f"Execution file not found: {plan}")
         sys.exit(1)
-    
     except ValueError as e:
         click.echo(f"Error: {e}", err=True)
-        logger.error(f"Execution failed: {e}")
+        ctx.logger.error(f"Execution validation failed: {e}")
         sys.exit(1)
-    
-    except Exception as e:
+    except OSError as e:
         click.echo(f"Error during execution: {e}", err=True)
-        logger.error(f"Execution failed: {e}", exc_info=True)
+        ctx.logger.error(f"Execution I/O failed: {e}", exc_info=True)
         sys.exit(1)
 
 
@@ -1177,114 +776,20 @@ def rollback(ctx: CLIContext, log: Optional[Path]):
         vlm rollback --log rollback_<uuid>.json         # Use specific log
         vlm rollback --log ~/.vlm/rollback/rollback_*.json
     """
-    from vlm.executor import ExecutionEngine
-    
-    config = ctx.config
-    logger = ctx.logger
-    
     try:
-        # If no log specified, find the most recent rollback log
-        if log is None:
-            rollback_dir = Path.home() / ".vlm" / "rollback"
-            if not rollback_dir.exists():
-                click.echo("Error: No rollback logs found.", err=True)
-                click.echo(f"Rollback directory does not exist: {rollback_dir}", err=True)
-                sys.exit(1)
-            
-            # Find all rollback log files
-            rollback_logs = sorted(rollback_dir.glob("rollback_*.json"), key=lambda p: p.stat().st_mtime, reverse=True)
-            
-            if not rollback_logs:
-                click.echo("Error: No rollback logs found.", err=True)
-                click.echo(f"No rollback_*.json files in: {rollback_dir}", err=True)
-                sys.exit(1)
-            
-            # Use the most recent log
-            log = rollback_logs[0]
-            click.echo(f"Using most recent rollback log: {log}")
-            click.echo()
-        
-        # Display rollback start message
-        click.echo(f"Loading rollback log from: {log}")
-        click.echo()
-        
-        # Create execution engine
-        engine = ExecutionEngine(logger=logger, config=config)
-        
-        # Load rollback log
-        rollback_log = engine.load_rollback_log(log)
-        
-        # Display rollback log info
-        click.echo(f"Rollback log loaded: {rollback_log.log_id}")
-        click.echo(f"Original execution: {rollback_log.execution_plan_id}")
-        click.echo(f"Executed at: {rollback_log.executed_at.strftime('%Y-%m-%d %H:%M:%S')}")
-        click.echo(f"Operations to rollback: {len(rollback_log.operations)}")
-        click.echo()
-        
-        # Display warning
-        click.echo("⚠️  ROLLBACK OPERATION - Best-effort restoration")
-        click.echo("   This will attempt to move files back to their original locations.")
-        click.echo("   Some operations may fail if files have been modified or moved.")
-        click.echo()
-        
-        if not click.confirm("Are you sure you want to proceed with rollback?"):
-            click.echo("Rollback cancelled.")
-            return
-        
-        click.echo()
-        click.echo(f"Rolling back {len(rollback_log.operations)} operations...")
-        click.echo()
-        
-        # Perform rollback
-        results, summary = engine.rollback(rollback_log)
-        
-        # Display rollback progress
-        for i, result in enumerate(results):
-            op = result.operation
-            if op.operation_type != "no-op":
-                status = "✓" if result.success else "✗"
-                click.echo(f"  [{i+1}/{len(results)}] {status} Rollback: {op.destination_path.name if op.destination_path else op.source_path.name}")
-                if result.error_message:
-                    click.echo(f"      Error: {result.error_message}")
-        
-        # Display rollback summary
-        click.echo()
-        click.echo("=" * 60)
-        click.echo("Rollback Summary")
-        click.echo("=" * 60)
-        click.echo(f"  Total operations: {summary['total']}")
-        click.echo(f"  Successful: {summary['successful']}")
-        click.echo(f"  Failed: {summary['failed']}")
-        click.echo(f"  Skipped: {summary['skipped']}")
-        click.echo()
-        
-        # Display completion message
-        if summary['failed'] > 0:
-            click.echo(f"⚠️  Rollback completed with {summary['failed']} failures.")
-            click.echo("   Check the log file for details.")
-            click.echo("   Some files may not have been restored to their original locations.")
-        else:
-            click.echo("✓ Rollback completed successfully!")
-            click.echo("  All files have been restored to their original locations.")
-        
-        logger.info(
-            f"Rollback completed: "
-            f"{summary['successful']} successful, {summary['failed']} failed, {summary['skipped']} skipped"
-        )
-        
+        from vlm.commands.execute import rollback_cmd
+        rollback_cmd(ctx, log)
     except FileNotFoundError as e:
         click.echo(f"Error: {e}", err=True)
-        logger.error(f"Rollback log not found: {e}")
+        ctx.logger.error(f"Rollback log not found: {e}")
         sys.exit(1)
-    
     except ValueError as e:
-        click.echo(f"Error: Invalid rollback log format: {e}", err=True)
-        logger.error(f"Invalid rollback log: {e}")
+        click.echo(f"Error: {e}", err=True)
+        ctx.logger.error(f"Rollback failed: {e}")
         sys.exit(1)
-    
-    except Exception as e:
+    except OSError as e:
         click.echo(f"Error during rollback: {e}", err=True)
-        logger.error(f"Rollback failed: {e}", exc_info=True)
+        ctx.logger.error(f"Rollback failed: {e}", exc_info=True)
         sys.exit(1)
 
 

src/vlm/commands/analyze.py

@@ -1,6 +1,5 @@
 """Analyze command implementation."""
 
-import json
 from pathlib import Path
 from typing import Optional
 
@@ -8,9 +7,13 @@ import click
 
 from vlm.analysis import analyze_series_completeness, detect_duplicates
 from vlm.context import CLIContext
-from vlm.io import identities_to_analysis_input, load_identities_json, load_inventory_csv
+from vlm.io import (
+    identities_to_analysis_input,
+    load_identities_json,
+    load_inventory_csv,
+    save_analysis_json,
+)
 from vlm.models import MovieIdentity, SeriesIdentity
-from vlm.utils import utc_now
 
 
 def analyze_cmd(
@@ -68,7 +71,6 @@ def analyze_cmd(
     click.echo(f"Saving analysis results to: {output}")
     output.parent.mkdir(parents=True, exist_ok=True)
 
-    generation_timestamp = utc_now().strftime("%Y-%m-%dT%H:%M:%S")
     completeness_list = [
         {
             "series_title": c.series_title,
@@ -96,19 +98,14 @@ def analyze_cmd(
                 "quality_comparison": d.quality_comparison,
             }
         )
-    analysis_data = {
-        "vlm_schema_version": "1.0",
-        "metadata": {
-            "generated": generation_timestamp,
-            "source_identities": str(input),
-            "total_movies": len(movies_data),
-            "total_series": len(series_data),
-        },
-        "completeness": completeness_list,
-        "duplicates": duplicates_list,
-    }
-    with open(output, "w", encoding="utf-8") as jsonfile:
-        json.dump(analysis_data, jsonfile, indent=2, ensure_ascii=False)
+    save_analysis_json(
+        completeness=completeness_list,
+        duplicates=duplicates_list,
+        source_identities=input,
+        total_movies=len(movies_data),
+        total_series=len(series_data),
+        output=output,
+    )
 
     click.echo("Analysis results saved successfully!")
     logger.info(

src/vlm/commands/enrich.py

@@ -0,0 +1,146 @@
+"""Enrich command implementation."""
+
+from __future__ import annotations
+
+import sys
+from pathlib import Path
+from typing import Optional
+
+import click
+
+from vlm.context import CLIContext
+from vlm.enrichment import enrich_identities_data
+from vlm.io import load_json_file, save_json_file
+
+
+def enrich_cmd(
+    ctx: CLIContext,
+    input: Path,
+    output: Optional[Path],
+    refresh_changed_only: bool,
+    refresh_all: bool,
+    timeout: int,
+    retries: int,
+) -> None:
+    """Enrich identities with translation and reputation metadata."""
+    logger = ctx.logger
+    config = ctx.config
+
+    if output is None:
+        output = input
+
+    click.echo(f"Enriching identities from: {input}")
+    click.echo(f"Output file: {output}")
+    click.echo()
+
+    identities_data = load_json_file(input)
+
+    if refresh_all and refresh_changed_only:
+        raise ValueError("--refresh-all and --refresh-changed-only are mutually exclusive.")
+    if timeout < 1:
+        raise ValueError("--timeout must be >= 1")
+    if retries < 0:
+        raise ValueError("--retries must be >= 0")
+
+    refresh_mode = "incremental"
+    if refresh_all:
+        refresh_mode = "refresh_all"
+    elif refresh_changed_only:
+        refresh_mode = "refresh_changed_only"
+
+    total = (
+        len(identities_data.get("movies", []))
+        + len(identities_data.get("series", []))
+        + len(identities_data.get("anime", []))
+    )
+
+    progress_state: dict[str, Optional[click.ProgressBar]] = {"bar": None}
+    progress_position = {"current": 0}
+    progress_bucket = {"value": -1}
+    is_tty = bool(getattr(sys.stderr, "isatty", lambda: False)())
+
+    def _enrich_progress(processed: int, total_count: int, metrics: dict[str, int]) -> None:
+        if total_count <= 0:
+            return
+
+        if is_tty and progress_state["bar"] is None:
+            bar = click.progressbar(
+                length=total_count,
+                label="Enriching records",
+                show_pos=True,
+            )
+            progress_state["bar"] = bar.__enter__()
+
+        step = processed - progress_position["current"]
+        if step > 0 and progress_state["bar"] is not None:
+            progress_state["bar"].update(step)
+            progress_position["current"] = processed
+
+        if not is_tty:
+            percent = int(processed * 100 / total_count)
+            bucket = percent // 5
+            if bucket > progress_bucket["value"] or processed == total_count:
+                progress_bucket["value"] = bucket
+                click.echo(
+                    "Progress: "
+                    f"{processed}/{total_count} ({percent}%) "
+                    f"api_calls={metrics.get('api_calls', 0)} "
+                    f"cache_hits={metrics.get('cache_hits', 0)} "
+                    f"failed={metrics.get('failed', 0)}"
+                )
+
+    try:
+        if total == 0:
+            click.echo("No movie/series/anime records found to enrich.")
+        enriched_data, stats = enrich_identities_data(
+            identities_data,
+            config,
+            refresh_mode=refresh_mode,
+            request_timeout=timeout,
+            retries=retries,
+            logger=logger,
+            progress_callback=_enrich_progress,
+        )
+    finally:
+        if progress_state["bar"] is not None:
+            progress_state["bar"].__exit__(None, None, None)
+            click.echo()
+
+    save_json_file(enriched_data, output)
+
+    total_records = int(stats["total"]) if stats["total"] else 0
+    cache_hits = int(stats["cache_hits"])
+    hit_rate = (cache_hits / total_records * 100.0) if total_records else 0.0
+
+    click.echo("Enrichment complete!")
+    click.echo(f"  Total records: {total_records}")
+    click.echo(f"  Refresh mode: {refresh_mode}")
+    click.echo(f"  Enriched now: {stats['enriched']}")
+    click.echo(f"  Cache hits: {stats['cache_hits']}")
+    click.echo(f"  Cache hit rate: {hit_rate:.1f}%")
+    click.echo(f"  API calls: {stats['api_calls']}")
+    click.echo(f"  Failed requests: {stats['failed']}")
+    click.echo(f"  Skipped: {stats['skipped']}")
+    click.echo(f"  Needs review: {stats['needs_review']}")
+
+    skip_reasons = stats.get("skip_reasons", {})
+    if isinstance(skip_reasons, dict):
+        non_zero = [f"{name}={count}" for name, count in sorted(skip_reasons.items()) if int(count) > 0]
+        if non_zero:
+            click.echo(f"  Skip reasons: {' '.join(non_zero)}")
+
+    failed_items = stats.get("failed_items", [])
+    if isinstance(failed_items, list) and failed_items:
+        click.echo("  Failure sample:")
+        for item in failed_items[:3]:
+            click.echo(f"    - [{item.get('provider', 'unknown')}] {item.get('title', '')}: {item.get('reason', '')}")
+
+    logger.info(
+        "Enrich completed: total=%s enriched=%s cache_hits=%s failed=%s skipped=%s mode=%s",
+        stats["total"],
+        stats["enriched"],
+        stats["cache_hits"],
+        stats["failed"],
+        stats["skipped"],
+        refresh_mode,
+    )

src/vlm/commands/execute.py

@@ -0,0 +1,257 @@
+"""Execute and rollback command implementations."""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Optional
+
+import click
+
+from vlm.context import CLIContext
+from vlm.executor import ExecutionEngine
+from vlm.planner import load_plan
+from vlm.state import StateManager
+
+
+def _validate_plan_structure(execution_plan) -> list[str]:
+    """Validate source/destination paths for move/rename operations."""
+    validation_errors: list[str] = []
+
+    for operation in execution_plan.operations:
+        if operation.operation_type in ("move", "rename"):
+            if not operation.source_path.exists():
+                validation_errors.append(f"Source file does not exist: {operation.source_path}")
+            elif not operation.source_path.is_file():
+                validation_errors.append(f"Source path is not a file: {operation.source_path}")
+
+    for operation in execution_plan.operations:
+        if operation.operation_type in ("move", "rename") and operation.destination_path:
+            dest_dir = operation.destination_path.parent
+            if not dest_dir.exists() and not dest_dir.parent.exists():
+                validation_errors.append(f"Cannot create destination directory (parent missing): {dest_dir}")
+
+    return validation_errors
+
+
+def execute_cmd(
+    ctx: CLIContext,
+    plan: Path,
+    confirm: bool,
+    yes: bool,
+    verbose_ops: bool,
+    preserve_directories: bool,
+    safe_mode: bool,
+) -> None:
+    """Execute an execution plan in dry-run or execute mode."""
+    config = ctx.config
+    logger = ctx.logger
+    mode = "execute" if confirm else "dry-run"
+
+    click.echo(f"Loading execution plan from: {plan}")
+    click.echo()
+
+    execution_plan = load_plan(plan)
+
+    if preserve_directories:
+        click.echo("Directory preservation is enabled (plan metadata/operations will be honored).")
+        click.echo()
+
+    if safe_mode:
+        click.echo("Safe mode enabled - validating plan for directory preservation...")
+        emptied_dirs = execution_plan.metadata.get("emptied_directories", [])
+        if emptied_dirs:
+            raise ValueError(
+                "SAFE MODE VIOLATION: plan would empty directories. "
+                "Regenerate/adjust the plan to preserve directory structure."
+            )
+        click.echo("Safe mode validation passed - no directories would be destroyed")
+        click.echo()
+
+    click.echo("Validating directory structure...")
+    validation_errors = _validate_plan_structure(execution_plan)
+    if validation_errors:
+        formatted = "\n".join(f"  - {error}" for error in validation_errors[:10])
+        if len(validation_errors) > 10:
+            formatted += f"\n  ... and {len(validation_errors) - 10} more errors"
+        raise ValueError(f"DIRECTORY STRUCTURE VALIDATION FAILED:\n{formatted}")
+
+    click.echo("Directory structure validation passed")
+    click.echo()
+
+    click.echo(f"Execution plan loaded: {execution_plan.plan_id}")
+    click.echo(f"Created at: {execution_plan.created_at}")
+    click.echo(f"Total operations: {len(execution_plan.operations)}")
+
+    if execution_plan.human_summary:
+        click.echo()
+        click.echo(execution_plan.human_summary)
+    elif execution_plan.summary or execution_plan.summary_by_reason:
+        s = execution_plan.summary or {}
+        by_r = execution_plan.summary_by_reason or {}
+        parts = [
+            "操作统计：共 "
+            f"{s.get('total', len(execution_plan.operations))} 条"
+            f"（move {s.get('move', 0)}，rename {s.get('rename', 0)}，"
+            f"quarantine {s.get('quarantine', 0)}，no-op {s.get('no-op', 0)}）"
+        ]
+        if by_r:
+            parts.append("原因分布：" + "；".join(f"{r}: {c}" for r, c in list(by_r.items())[:5]))
+        click.echo()
+        click.echo("\n".join(parts))
+    click.echo()
+
+    if mode == "dry-run":
+        click.echo("DRY-RUN MODE - No files will be modified")
+        click.echo("Use --confirm to actually execute operations")
+    else:
+        click.echo("EXECUTE MODE - Files will be modified")
+        click.echo()
+        if not yes and not click.confirm("Are you sure you want to proceed?"):
+            click.echo("Execution cancelled.")
+            return
+        if yes:
+            click.echo("Auto-approved via --yes flag")
+
+    click.echo()
+    click.echo(f"Executing {len(execution_plan.operations)} operations...")
+    click.echo()
+
+    state_path = Path.home() / ".vlm" / "state.json"
+    state_manager = StateManager(state_path)
+
+    engine = ExecutionEngine(
+        logger=logger,
+        config=config,
+        verbose_operations=verbose_ops,
+        state_manager=state_manager,
+    )
+    results, summary, rollback_log = engine.execute_plan(
+        execution_plan,
+        mode=mode,
+        confirmed=confirm,
+    )
+
+    if mode == "dry-run":
+        click.echo("Sample operations (dry-run):")
+        for i, result in enumerate(results[:5]):
+            op = result.operation
+            if op.operation_type != "no-op":
+                click.echo(f"  [{i + 1}] {op.operation_type}: {op.source_path.name}")
+                if op.destination_path:
+                    click.echo(f"      -> {op.destination_path}")
+        if len(results) > 5:
+            click.echo(f"  ... and {len(results) - 5} more operations")
+    else:
+        for i, result in enumerate(results):
+            op = result.operation
+            if op.operation_type != "no-op" and not op.has_conflict:
+                status = "OK" if result.success else "FAIL"
+                click.echo(f"  [{i + 1}/{len(results)}] {status} {op.operation_type}: {op.source_path.name}")
+                if result.error_message:
+                    click.echo(f"      Error: {result.error_message}")
+
+    click.echo()
+    click.echo("=" * 60)
+    click.echo(f"Execution Summary ({mode} mode)")
+    click.echo("=" * 60)
+    click.echo(f"  Total operations: {summary['total']}")
+    click.echo(f"  Successful: {summary['successful']}")
+    click.echo(f"  Failed: {summary['failed']}")
+    click.echo(f"  Skipped: {summary['skipped']}")
+    click.echo()
+
+    if mode == "execute" and rollback_log:
+        rollback_dir = Path.home() / ".vlm" / "rollback"
+        rollback_dir.mkdir(parents=True, exist_ok=True)
+        rollback_path = rollback_dir / f"rollback_{rollback_log.log_id}.json"
+        engine.save_rollback_log(rollback_log, rollback_path)
+
+        click.echo(f"Rollback log saved to: {rollback_path}")
+        click.echo()
+        click.echo("To undo these operations, run:")
+        click.echo(f"  vlm rollback --log {rollback_path}")
+        click.echo()
+
+    if mode == "dry-run":
+        click.echo("Dry-run complete! No files were modified.")
+        click.echo("Review the operations above and use --confirm to execute.")
+    elif summary["failed"] > 0:
+        click.echo(f"Execution completed with {summary['failed']} failures.")
+        click.echo("Check the log file for details.")
+    else:
+        click.echo("Execution completed successfully!")
+
+    logger.info(
+        "Execution completed in %s mode: %s successful, %s failed, %s skipped",
+        mode,
+        summary["successful"],
+        summary["failed"],
+        summary["skipped"],
+    )
+
+
+def rollback_cmd(ctx: CLIContext, log: Optional[Path]) -> None:
+    """Rollback a prior execution from rollback log."""
+    config = ctx.config
+    logger = ctx.logger
+
+    if log is None:
+        rollback_dir = Path.home() / ".vlm" / "rollback"
+        if not rollback_dir.exists():
+            raise FileNotFoundError(f"No rollback logs found. Rollback directory does not exist: {rollback_dir}")
+        rollback_logs = sorted(
+            rollback_dir.glob("rollback_*.json"), key=lambda p: p.stat().st_mtime, reverse=True
+        )
+        if not rollback_logs:
+            raise FileNotFoundError(f"No rollback logs found. No rollback_*.json files in: {rollback_dir}")
+        log = rollback_logs[0]
+
+    click.echo(f"Loading rollback log: {log}")
+
+    engine = ExecutionEngine(logger=logger, config=config)
+    rollback_log = engine.load_rollback_log(log)
+
+    click.echo(f"Rollback log loaded: {rollback_log.log_id}")
+    click.echo(f"Plan ID: {rollback_log.execution_plan_id}")
+    click.echo(f"Executed at: {rollback_log.executed_at}")
+    click.echo(f"Operations to rollback: {len(rollback_log.operations)}")
+    click.echo()
+
+    if not click.confirm("Proceed with rollback?"):
+        click.echo("Rollback cancelled.")
+        return
+
+    click.echo(f"Rolling back {len(rollback_log.operations)} operations...")
+    click.echo()
+    click.echo("Performing rollback...")
+    rollback_results, rollback_summary = engine.rollback(rollback_log)
+
+    click.echo()
+    click.echo("Rollback Summary")
+    click.echo("=" * 60)
+    click.echo(f"  Total operations: {rollback_summary['total']}")
+    click.echo(f"  Successful: {rollback_summary['successful']}")
+    click.echo(f"  Failed: {rollback_summary['failed']}")
+    click.echo(f"  Skipped: {rollback_summary['skipped']}")
+    click.echo()
+
+    if rollback_summary["failed"] > 0:
+        click.echo("Failed operations:")
+        for result in rollback_results:
+            if not result.success:
+                op = result.operation
+                click.echo(f"  - {op.operation_type}: {op.source_path}")
+                if result.error_message:
+                    click.echo(f"    Error: {result.error_message}")
+        click.echo()
+
+    if rollback_summary["failed"] == 0:
+        click.echo("Rollback completed successfully!")
+    else:
+        click.echo(f"Rollback completed with {rollback_summary['failed']} failures.")
+
+    logger.info(
+        "Rollback completed: %s successful, %s failed",
+        rollback_summary["successful"],
+        rollback_summary["failed"],
+    )

src/vlm/commands/parse.py

@@ -0,0 +1,165 @@
+"""Parse command implementation."""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Optional
+
+import click
+
+from vlm.context import CLIContext
+from vlm.io import load_inventory_csv, save_identities_json
+from vlm.parser import parse_movie, parse_series
+from vlm.utils import utc_now
+
+
+def parse_cmd(ctx: CLIContext, input: Path, output: Path, inventory: Optional[Path]) -> None:
+    """Parse identities from scanned inventory."""
+    config = ctx.config
+    logger = ctx.logger
+
+    click.echo(f"Parsing identities from: {input}")
+
+    path_to_metadata: dict[str, object] = {}
+    if inventory:
+        click.echo(f"Loading video metadata from: {inventory}")
+        inventory_files = load_inventory_csv(inventory)
+        path_to_metadata = {str(vf.path): vf for vf in inventory_files}
+        click.echo(f"Loaded metadata for {len(path_to_metadata)} files")
+
+    click.echo()
+
+    inventory_files = load_inventory_csv(input)
+    video_files = [
+        {
+            "path": str(vf.path),
+            "filename": vf.filename,
+            "category": vf.category,
+        }
+        for vf in inventory_files
+    ]
+
+    click.echo(f"Loaded {len(video_files)} files from inventory")
+    click.echo()
+
+    movie_identities: list[dict] = []
+    series_identities: list[dict] = []
+    anime_files: list[dict] = []
+    other_files: list[dict] = []
+
+    def get_video_metadata(file_path: str) -> dict:
+        """Extract video metadata from inventory if available."""
+        if not path_to_metadata:
+            return {}
+        vf = path_to_metadata.get(file_path)
+        if not vf:
+            return {}
+        return {
+            "size_bytes": vf.size_bytes,
+            "modified_timestamp": vf.modified_timestamp.isoformat(),
+            "resolution": vf.resolution,
+            "codec": vf.codec,
+            "duration_seconds": vf.duration_seconds,
+            "bitrate_kbps": vf.bitrate_kbps,
+        }
+
+    for vf in video_files:
+        filename = vf["filename"]
+        category = vf["category"]
+        file_path = vf["path"]
+        video_metadata = get_video_metadata(file_path)
+
+        if category == "movie":
+            identity = parse_movie(filename, extensions=config.video_extensions)
+            record = {
+                "path": file_path,
+                "filename": filename,
+                "category": category,
+                "title": identity.title,
+                "year": identity.year,
+                "confidence": identity.confidence,
+                "needs_review": identity.needs_review,
+            }
+            if video_metadata:
+                record["video_metadata"] = video_metadata
+            movie_identities.append(record)
+        elif category == "series":
+            identity = parse_series(filename, extensions=config.video_extensions)
+            record = {
+                "path": file_path,
+                "filename": filename,
+                "category": category,
+                "title": identity.title,
+                "season": identity.season,
+                "episodes": identity.episodes,
+                "confidence": identity.confidence,
+                "needs_review": identity.needs_review,
+            }
+            if video_metadata:
+                record["video_metadata"] = video_metadata
+            series_identities.append(record)
+        elif category == "anime":
+            anime_files.append(
+                {
+                    "path": vf["path"],
+                    "filename": filename,
+                    "category": category,
+                    "note": "Anime parsing deferred in v1",
+                }
+            )
+        else:
+            other_files.append(
+                {
+                    "path": vf["path"],
+                    "filename": filename,
+                    "category": category,
+                    "note": "Not categorized for parsing",
+                }
+            )
+
+    click.echo("Parsing complete!")
+    click.echo()
+    click.echo("Results by category:")
+    click.echo(f"  Movies: {len(movie_identities)}")
+
+    movies_need_review = sum(1 for m in movie_identities if m["needs_review"])
+    if movies_need_review > 0:
+        click.echo(f"    - Need review: {movies_need_review}")
+
+    click.echo(f"  Series: {len(series_identities)}")
+
+    series_need_review = sum(1 for s in series_identities if s["needs_review"])
+    if series_need_review > 0:
+        click.echo(f"    - Need review: {series_need_review}")
+
+    click.echo(f"  Anime: {len(anime_files)} (not parsed in v1)")
+    click.echo(f"  Other: {len(other_files)} (not parsed)")
+
+    click.echo()
+    click.echo(f"Saving parsed identities to: {output}")
+
+    generation_timestamp = utc_now().strftime("%Y-%m-%dT%H:%M:%S")
+    schema_version = "2.0" if path_to_metadata else "1.0"
+
+    identities_data = {
+        "vlm_schema_version": schema_version,
+        "metadata": {
+            "generated": generation_timestamp,
+            "source_inventory": str(input),
+            "total_files": len(video_files),
+        },
+        "movies": movie_identities,
+        "series": series_identities,
+        "anime": anime_files,
+        "other": other_files,
+    }
+
+    save_identities_json(identities_data, output)
+
+    click.echo("Parsed identities saved successfully!")
+    logger.info(
+        "Parse completed: %s movies, %s series, saved to %s",
+        len(movie_identities),
+        len(series_identities),
+        output,
+    )

src/vlm/commands/plan.py

@@ -81,6 +81,15 @@ def plan_cmd(ctx: CLIContext, input: Path, output: Path, analysis: Optional[Path
         click.echo(f"  ⚠️  Conflicts detected: {conflicts}")
         click.echo("     Review the plan file for details on conflicting operations.")
 
+    # Check for directory warnings
+    directory_warning = execution_plan.metadata.get("directory_warning", False)
+    if directory_warning:
+        emptied_dirs = execution_plan.metadata.get("emptied_directories", [])
+        click.echo()
+        click.echo(f"  ⚠️  Directory preservation warning: {len(emptied_dirs)} directories will be emptied")
+        click.echo("     These directories will be preserved but may be empty after execution.")
+        click.echo("     Review the plan file for details on preserved directories.")
+
     click.echo()
     click.echo(f"Saving execution plan to: {output}")
     output.parent.mkdir(parents=True, exist_ok=True)

src/vlm/config.py

@@ -5,15 +5,17 @@ from pathlib import Path
 from typing import Optional
 import yaml
 
+DEFAULT_VIDEO_EXTENSIONS = [
+    ".mp4", ".mkv", ".avi", ".mov", ".wmv", ".flv", ".webm", ".m4v"
+]
+
 
 @dataclass
 class Config:
     """Configuration for Video Library Manager."""
 
     library_root: Path
-    video_extensions: list[str] = field(default_factory=lambda: [
-        ".mp4", ".mkv", ".avi", ".mov", ".wmv", ".flv", ".webm", ".m4v"
-    ])
+    video_extensions: list[str] = field(default_factory=lambda: list(DEFAULT_VIDEO_EXTENSIONS))
     movie_template: str = "movie/{title} ({year})/"
     series_template: str = "series/{title}/Season {season:02d}/"
     movie_filename_template: str = "{title} ({year}){ext}"
@@ -74,9 +76,7 @@ def load_config(path: Path) -> Config:
 
     library_root = Path(library_root_str).expanduser()
 
-    video_extensions = data.get("video_extensions", [
-        ".mp4", ".mkv", ".avi", ".mov", ".wmv", ".flv", ".webm", ".m4v"
-    ])
+    video_extensions = data.get("video_extensions", list(DEFAULT_VIDEO_EXTENSIONS))
 
     templates = data.get("templates", {})
     movie_template = templates.get("movie_dir", "movie/{title} ({year})/")
@@ -149,7 +149,7 @@ def create_default_config(path: Path) -> Config:
     """Create a default configuration file and return the Config object."""
     default_config = Config(
         library_root=Path.home() / "Videos",
-        video_extensions=[".mp4", ".mkv", ".avi", ".mov", ".wmv", ".flv", ".webm", ".m4v"],
+        video_extensions=list(DEFAULT_VIDEO_EXTENSIONS),
     )
 
     enrichment_content = {

src/vlm/executor.py

@@ -93,8 +93,17 @@ class ExecutionEngine:
         transaction_log = None
         if mode == "execute":
             log_path = Path.home() / ".vlm" / "transaction.json"
+            try:
                 transaction_log = TransactionLog(log_path)
                 transaction_log.start_transaction(plan)
+            except OSError as exc:
+                log_operation(
+                    self.logger,
+                    logging.WARNING,
+                    f"Transaction log disabled (cannot write {log_path}): {exc}",
+                    operation_type="execute",
+                )
+                transaction_log = None
         
         # Execute all operations
         results = []
@@ -105,9 +114,18 @@ class ExecutionEngine:
             # Update transaction and state logs in execute mode
             if mode == "execute":
                 if transaction_log:
+                    try:
                         transaction_log.mark_operation_complete(
                             i, result.success, result.error_message
                         )
+                    except OSError as exc:
+                        log_operation(
+                            self.logger,
+                            logging.WARNING,
+                            f"Failed to update transaction log: {exc}",
+                            operation_type="execute",
+                        )
+                        transaction_log = None
                 
                 # Update file state if successful and not a no-op
                 if result.success and operation.operation_type != "no-op" and self.state_manager:
@@ -126,7 +144,15 @@ class ExecutionEngine:
         if mode == "execute":
             if transaction_log:
                 status = "completed" if all(r.success for r in results) else "failed"
+                try:
                     transaction_log.complete_transaction(status=status)
+                except OSError as exc:
+                    log_operation(
+                        self.logger,
+                        logging.WARNING,
+                        f"Failed to finalize transaction log: {exc}",
+                        operation_type="execute",
+                    )
             if self.state_manager:
                 self.state_manager.save()
         
@@ -189,6 +215,22 @@ class ExecutionEngine:
                 executed_at=executed_at
             )
 
+        # Handle preserve-directory operations
+        if operation.operation_type == "preserve-directory":
+            log_operation(
+                self.logger,
+                logging.DEBUG,
+                f"Preserving directory: {operation.reason}",
+                operation_type="execute",
+                file_path=operation.source_path
+            )
+            return OperationResult(
+                operation=operation,
+                success=True,
+                error_message=None,
+                executed_at=executed_at
+            )
+
         # Handle quarantine operations (no destination_path; use QuarantineManager)
         if operation.operation_type == "quarantine":
             if not self._quarantine_manager:

src/vlm/io.py

@@ -15,34 +15,71 @@ from vlm.scanner import load_inventory_csv, save_inventory_csv
 __all__ = [
     "load_inventory_csv",
     "save_inventory_csv",
+    "load_json_file",
+    "save_json_file",
     "load_identities_json",
     "save_identities_json",
     "load_analysis_json",
+    "save_analysis_json",
     "identities_to_plan_input",
     "identities_to_analysis_input",
 ]
 
 
+def load_json_file(path: Path) -> dict:
+    """Load a JSON object from disk."""
+    with open(path, "r", encoding="utf-8") as f:
+        return json.load(f)
+
+
+def save_json_file(data: dict, path: Path) -> None:
+    """Save a JSON object to disk."""
+    path.parent.mkdir(parents=True, exist_ok=True)
+    with open(path, "w", encoding="utf-8") as f:
+        json.dump(data, f, indent=2, ensure_ascii=False)
+
+
 def load_analysis_json(path: Path) -> dict:
     """Load analysis result from JSON file (metadata, completeness, duplicates).
 
     Caller should check file existence and handle missing/invalid keys.
     """
-    with open(path, "r", encoding="utf-8") as f:
-        return json.load(f)
+    return load_json_file(path)
 
 
 def load_identities_json(path: Path) -> dict:
     """Load identities from JSON file."""
-    with open(path, "r", encoding="utf-8") as f:
-        return json.load(f)
+    return load_json_file(path)
 
 
 def save_identities_json(data: dict, path: Path) -> None:
     """Save identities dict to JSON file."""
-    path.parent.mkdir(parents=True, exist_ok=True)
-    with open(path, "w", encoding="utf-8") as f:
-        json.dump(data, f, indent=2, ensure_ascii=False)
+    save_json_file(data, path)
+
+
+def save_analysis_json(
+    *,
+    completeness: list[dict],
+    duplicates: list[dict],
+    source_identities: Path,
+    total_movies: int,
+    total_series: int,
+    output: Path,
+) -> None:
+    """Save analysis result JSON using the canonical schema."""
+    generation_timestamp = utc_now().strftime("%Y-%m-%dT%H:%M:%S")
+    analysis_data = {
+        "vlm_schema_version": "1.0",
+        "metadata": {
+            "generated": generation_timestamp,
+            "source_identities": str(source_identities),
+            "total_movies": total_movies,
+            "total_series": total_series,
+        },
+        "completeness": completeness,
+        "duplicates": duplicates,
+    }
+    save_json_file(analysis_data, output)
 
 
 def _video_file_from_record(record: dict) -> VideoFile:

src/vlm/models.py

@@ -115,7 +115,7 @@ class FileOperation:
     """Represents a single file operation in an execution plan.
 
     Attributes:
-        operation_type: Type of operation ("move", "rename", "quarantine", "no-op")
+        operation_type: Type of operation ("move", "rename", "quarantine", "no-op", "preserve-directory")
         source_path: Source file path
         destination_path: Destination file path (None for no-op operations)
         reason: Human-readable reason for the operation

src/vlm/parser.py

@@ -7,14 +7,9 @@ logical identities such as movie titles/years and series titles/seasons/episodes
 import re
 from typing import Optional
 
+from vlm.config import DEFAULT_VIDEO_EXTENSIONS
 from vlm.models import MovieIdentity, SeriesIdentity
 
-# Default extensions used when extensions param is not provided (matches config default)
-DEFAULT_VIDEO_EXTENSIONS = [
-    ".mp4", ".mkv", ".avi", ".mov", ".wmv", ".flv", ".webm", ".m4v"
-]
-
-
 # Quality tags to remove from titles
 QUALITY_TAGS = [
     r'\b1080p\b', r'\b720p\b', r'\b480p\b', r'\b2160p\b',

src/vlm/planner.py

@@ -40,6 +40,42 @@ def _is_sample_path(path: Path) -> bool:
     return bool(re.search(r"(^|[\s._-])sample($|[\s._-])", stem))
 
 
+def _analyze_directory_impact(operations: list[FileOperation]) -> dict:
+    """Analyze which directories will be emptied by the plan."""
+    # Get all source directories that have files being moved/renamed
+    source_dirs = set()
+    moved_files = set()
+    for operation in operations:
+        if operation.operation_type in ("move", "rename"):
+            source_dirs.add(operation.source_path.parent)
+            moved_files.add(operation.source_path)
+
+    emptied_dirs = []
+    for dir_path in source_dirs:
+        # Only analyze directories that actually exist
+        if not dir_path.exists():
+            continue
+
+        # Count files that will remain in this directory after operations
+        remaining_count = 0
+        try:
+            for item in dir_path.iterdir():
+                if item.is_file() and item not in moved_files:
+                    remaining_count += 1
+        except (OSError, PermissionError):
+            # If we can't read the directory, skip analysis
+            continue
+
+        # If no files will remain, this directory will be emptied
+        if remaining_count == 0:
+            emptied_dirs.append(dir_path)
+
+    return {
+        "emptied_directories": emptied_dirs,
+        "warning_required": len(emptied_dirs) > 0
+    }
+
+
 def generate_plan(
     identities: list[tuple[VideoFile, Union[MovieIdentity, SeriesIdentity, None]]],
     config: Config,
@@ -120,10 +156,29 @@ def generate_plan(
                         conflict_reason=None,
                     )
 
+    # Analyze directory impact and add preservation operations
+    directory_analysis = _analyze_directory_impact(operations)
+    if directory_analysis["warning_required"]:
+        # Add directory preservation operations for emptied directories
+        for emptied_dir in directory_analysis["emptied_directories"]:
+            operations.append(FileOperation(
+                operation_type="preserve-directory",
+                source_path=emptied_dir,
+                destination_path=None,
+                reason=f"Preserve empty source directory: {emptied_dir.name}",
+                has_conflict=False,
+                conflict_reason=None
+            ))
+
     summary = _generate_summary(operations)
     summary_by_reason = _generate_summary_by_reason(operations)
     human_summary = _generate_human_summary(operations, summary, summary_by_reason, metadata)
 
+    # Add directory warnings to metadata
+    if directory_analysis["warning_required"]:
+        metadata["emptied_directories"] = [str(d) for d in directory_analysis["emptied_directories"]]
+        metadata["directory_warning"] = True
+
     return ExecutionPlan(
         plan_id=str(uuid.uuid4()),
         created_at=utc_now(),
@@ -445,7 +500,8 @@ def _generate_summary(operations: list[FileOperation]) -> dict:
         "move": 0,
         "rename": 0,
         "quarantine": 0,
-        "no-op": 0
+        "no-op": 0,
+        "preserve-directory": 0
     }
 
     for operation in operations:

tests/test_analysis.py

@@ -5,7 +5,7 @@ Tests series completeness analysis, duplicate detection, and quality comparison.
 
 import pytest
 from pathlib import Path
-from datetime import datetime
+from datetime import datetime, timezone
 from vlm.models import SeriesIdentity, SeasonCompleteness, MovieIdentity, VideoFile, DuplicateGroup
 from vlm.analysis import analyze_series_completeness, detect_duplicates, compare_quality
 
@@ -193,7 +193,7 @@ class TestDuplicateDetection:
                 Path("/movies/The.Matrix.1999.1080p.mkv"),
                 "The.Matrix.1999.1080p.mkv",
                 2000000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie",
                 resolution="1920x1080",
                 codec="h264"
@@ -202,7 +202,7 @@ class TestDuplicateDetection:
                 Path("/movies/The.Matrix.1999.720p.mkv"),
                 "The.Matrix.1999.720p.mkv",
                 1000000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie",
                 resolution="1280x720",
                 codec="h264"
@@ -211,7 +211,7 @@ class TestDuplicateDetection:
                 Path("/movies/Inception.2010.mkv"),
                 "Inception.2010.mkv",
                 1500000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie"
             ),
         ]
@@ -239,7 +239,7 @@ class TestDuplicateDetection:
                 Path("/series/Breaking.Bad.S01E01.1080p.mkv"),
                 "Breaking.Bad.S01E01.1080p.mkv",
                 1500000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "series",
                 resolution="1920x1080"
             ),
@@ -247,7 +247,7 @@ class TestDuplicateDetection:
                 Path("/series/Breaking.Bad.S01E01.720p.mkv"),
                 "Breaking.Bad.S01E01.720p.mkv",
                 800000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "series",
                 resolution="1280x720"
             ),
@@ -255,7 +255,7 @@ class TestDuplicateDetection:
                 Path("/series/Breaking.Bad.S01E02.mkv"),
                 "Breaking.Bad.S01E02.mkv",
                 1200000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "series"
             ),
         ]
@@ -278,8 +278,8 @@ class TestDuplicateDetection:
         ]
         
         files = [
-            VideoFile(Path("/movies/Movie.A.2020.mkv"), "Movie.A.2020.mkv", 1000000000, datetime.now(), "movie"),
-            VideoFile(Path("/movies/Movie.B.2021.mkv"), "Movie.B.2021.mkv", 1000000000, datetime.now(), "movie"),
+            VideoFile(Path("/movies/Movie.A.2020.mkv"), "Movie.A.2020.mkv", 1000000000, datetime.now(timezone.utc), "movie"),
+            VideoFile(Path("/movies/Movie.B.2021.mkv"), "Movie.B.2021.mkv", 1000000000, datetime.now(timezone.utc), "movie"),
         ]
         
         result = detect_duplicates(list(zip(identities, files)))
@@ -294,8 +294,8 @@ class TestDuplicateDetection:
         ]
         
         files = [
-            VideoFile(Path("/movies/Unknown.Movie.mkv"), "Unknown.Movie.mkv", 1000000000, datetime.now(), "movie"),
-            VideoFile(Path("/movies/Unknown.Movie.2.mkv"), "Unknown.Movie.2.mkv", 1000000000, datetime.now(), "movie"),
+            VideoFile(Path("/movies/Unknown.Movie.mkv"), "Unknown.Movie.mkv", 1000000000, datetime.now(timezone.utc), "movie"),
+            VideoFile(Path("/movies/Unknown.Movie.2.mkv"), "Unknown.Movie.2.mkv", 1000000000, datetime.now(timezone.utc), "movie"),
         ]
         
         result = detect_duplicates(list(zip(identities, files)))
@@ -311,8 +311,8 @@ class TestDuplicateDetection:
         ]
         
         files = [
-            VideoFile(Path("/series/Unknown.Show.E01.mkv"), "Unknown.Show.E01.mkv", 1000000000, datetime.now(), "series"),
-            VideoFile(Path("/series/Unknown.Show.Episode.1.mkv"), "Unknown.Show.Episode.1.mkv", 1000000000, datetime.now(), "series"),
+            VideoFile(Path("/series/Unknown.Show.E01.mkv"), "Unknown.Show.E01.mkv", 1000000000, datetime.now(timezone.utc), "series"),
+            VideoFile(Path("/series/Unknown.Show.Episode.1.mkv"), "Unknown.Show.Episode.1.mkv", 1000000000, datetime.now(timezone.utc), "series"),
         ]
         
         result = detect_duplicates(list(zip(identities, files)))
@@ -327,8 +327,8 @@ class TestDuplicateDetection:
         ]
         
         files = [
-            VideoFile(Path("/series/Show.Name.S01.mkv"), "Show.Name.S01.mkv", 1000000000, datetime.now(), "series"),
-            VideoFile(Path("/series/Show.Name.Season.1.mkv"), "Show.Name.Season.1.mkv", 1000000000, datetime.now(), "series"),
+            VideoFile(Path("/series/Show.Name.S01.mkv"), "Show.Name.S01.mkv", 1000000000, datetime.now(timezone.utc), "series"),
+            VideoFile(Path("/series/Show.Name.Season.1.mkv"), "Show.Name.Season.1.mkv", 1000000000, datetime.now(timezone.utc), "series"),
         ]
         
         result = detect_duplicates(list(zip(identities, files)))
@@ -347,7 +347,7 @@ class TestDuplicateDetection:
                 Path("/movies/Test.Movie.2020.1080p.mkv"),
                 "Test.Movie.2020.1080p.mkv",
                 2000000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie",
                 resolution="1920x1080",
                 codec="h264",
@@ -358,7 +358,7 @@ class TestDuplicateDetection:
                 Path("/movies/Test.Movie.2020.720p.mkv"),
                 "Test.Movie.2020.720p.mkv",
                 1000000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie",
                 resolution="1280x720",
                 codec="h264",
@@ -395,9 +395,9 @@ class TestDuplicateDetection:
         ]
         
         files = [
-            VideoFile(Path("/series/Show.S01E01-E02.mkv"), "Show.S01E01-E02.mkv", 2000000000, datetime.now(), "series"),
-            VideoFile(Path("/series/Show.S01E01.mkv"), "Show.S01E01.mkv", 1000000000, datetime.now(), "series"),
-            VideoFile(Path("/series/Show.S01E02.mkv"), "Show.S01E02.mkv", 1000000000, datetime.now(), "series"),
+            VideoFile(Path("/series/Show.S01E01-E02.mkv"), "Show.S01E01-E02.mkv", 2000000000, datetime.now(timezone.utc), "series"),
+            VideoFile(Path("/series/Show.S01E01.mkv"), "Show.S01E01.mkv", 1000000000, datetime.now(timezone.utc), "series"),
+            VideoFile(Path("/series/Show.S01E02.mkv"), "Show.S01E02.mkv", 1000000000, datetime.now(timezone.utc), "series"),
         ]
         
         result = detect_duplicates(list(zip(identities, files)))
@@ -413,8 +413,8 @@ class TestDuplicateDetection:
         ]
         
         files = [
-            VideoFile(Path("/movies/The.Thing.1982.mkv"), "The.Thing.1982.mkv", 1000000000, datetime.now(), "movie"),
-            VideoFile(Path("/movies/The.Thing.2011.mkv"), "The.Thing.2011.mkv", 1000000000, datetime.now(), "movie"),
+            VideoFile(Path("/movies/The.Thing.1982.mkv"), "The.Thing.1982.mkv", 1000000000, datetime.now(timezone.utc), "movie"),
+            VideoFile(Path("/movies/The.Thing.2011.mkv"), "The.Thing.2011.mkv", 1000000000, datetime.now(timezone.utc), "movie"),
         ]
         
         result = detect_duplicates(list(zip(identities, files)))
@@ -429,8 +429,8 @@ class TestDuplicateDetection:
         ]
         
         files = [
-            VideoFile(Path("/series/Show.S01E01.mkv"), "Show.S01E01.mkv", 1000000000, datetime.now(), "series"),
-            VideoFile(Path("/series/Show.S02E01.mkv"), "Show.S02E01.mkv", 1000000000, datetime.now(), "series"),
+            VideoFile(Path("/series/Show.S01E01.mkv"), "Show.S01E01.mkv", 1000000000, datetime.now(timezone.utc), "series"),
+            VideoFile(Path("/series/Show.S02E01.mkv"), "Show.S02E01.mkv", 1000000000, datetime.now(timezone.utc), "series"),
         ]
         
         result = detect_duplicates(list(zip(identities, files)))
@@ -448,7 +448,7 @@ class TestQualityComparison:
                 Path("/test/file1.mkv"),
                 "file1.mkv",
                 2000000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie",
                 resolution="1920x1080",
                 codec="h264",
@@ -459,7 +459,7 @@ class TestQualityComparison:
                 Path("/test/file2.mkv"),
                 "file2.mkv",
                 1000000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie",
                 resolution="1280x720",
                 codec="h265",
@@ -490,7 +490,7 @@ class TestQualityComparison:
                 Path("/test/file1.mkv"),
                 "file1.mkv",
                 2000000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie",
                 resolution="1920x1080"
                 # codec, duration, bitrate not available
@@ -499,7 +499,7 @@ class TestQualityComparison:
                 Path("/test/file2.mkv"),
                 "file2.mkv",
                 1000000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie"
                 # No optional metadata
             ),
@@ -532,7 +532,7 @@ class TestQualityComparison:
                 Path("/test/file.mkv"),
                 "file.mkv",
                 1500000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie",
                 resolution="1920x1080",
                 codec="h264"

tests/test_analysis_properties.py

@@ -6,7 +6,7 @@ Each test validates a specific property from the design document.
 
 import pytest
 from pathlib import Path
-from datetime import datetime
+from datetime import datetime, timezone
 from hypothesis import given, strategies as st, settings
 from vlm.models import (
     SeriesIdentity, SeasonCompleteness, MovieIdentity, VideoFile, DuplicateGroup
@@ -74,7 +74,7 @@ def video_file_strategy(draw, filename=None, category="movie"):
     
     path = Path(f"/{category}/{filename}")
     size_bytes = draw(st.integers(min_value=1000000, max_value=10000000000))
-    modified_timestamp = datetime.now()
+    modified_timestamp = datetime.now(timezone.utc)
     
     # Optional metadata
     has_metadata = draw(st.booleans())
@@ -226,7 +226,7 @@ def test_property_12_duplicate_detection_movies(title, year, duplicate_count):
             Path(f"/movies/{filename}"),
             filename,
             1000000000 + i * 100000000,
-            datetime.now(),
+            datetime.now(timezone.utc),
             "movie"
         ))
     
@@ -272,7 +272,7 @@ def test_property_13_duplicate_detection_series(title, season, episode, duplicat
             Path(f"/series/{filename}"),
             filename,
             1000000000 + i * 100000000,
-            datetime.now(),
+            datetime.now(timezone.utc),
             "series"
         ))
     
@@ -321,7 +321,7 @@ def test_property_14_duplicate_quality_comparison(title, year, file_count):
                 Path(f"/movies/{filename}"),
                 filename,
                 1000000000 + i * 100000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie",
                 resolution="1920x1080",
                 codec="h264",
@@ -333,7 +333,7 @@ def test_property_14_duplicate_quality_comparison(title, year, file_count):
                 Path(f"/movies/{filename}"),
                 filename,
                 1000000000 + i * 100000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie"
             ))
     
@@ -429,7 +429,7 @@ def test_property_43_duplicate_report_grouping(duplicate_count, format):
                 Path(f"/movies/{filename}"),
                 filename,
                 1000000000 + j * 500000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie",
                 resolution="1920x1080" if j == 0 else "1280x720",
                 codec="h264"
@@ -492,7 +492,7 @@ def test_property_44_summary_report_accuracy(file_count, categories):
             Path(f"/{category}/{filename}"),
             filename,
             size,
-            datetime.now(),
+            datetime.now(timezone.utc),
             category
         ))
         

tests/test_cli_rollback.py

@@ -2,7 +2,7 @@
 
 import json
 from pathlib import Path
-from datetime import datetime
+from datetime import datetime, timezone
 
 import pytest
 from click.testing import CliRunner
@@ -46,7 +46,7 @@ def sample_rollback_log(tmp_path):
             ),
             success=True,
             error_message=None,
-            executed_at=datetime.now()
+            executed_at=datetime.now(timezone.utc)
         ),
         OperationResult(
             operation=FileOperation(
@@ -59,14 +59,14 @@ def sample_rollback_log(tmp_path):
             ),
             success=True,
             error_message=None,
-            executed_at=datetime.now()
+            executed_at=datetime.now(timezone.utc)
         )
     ]
     
     rollback_log = RollbackLog(
         log_id="test-log-id",
         execution_plan_id="test-plan-id",
-        executed_at=datetime.now(),
+        executed_at=datetime.now(timezone.utc),
         operations=operations
     )
     

tests/test_executor.py

@@ -4,7 +4,7 @@ Tests execution mode handling, dry-run simulation, and actual file operations.
 """
 
 import logging
-from datetime import datetime
+from datetime import datetime, timezone
 from pathlib import Path
 from uuid import uuid4
 
@@ -73,7 +73,7 @@ def sample_plan(temp_test_dir):
     
     return ExecutionPlan(
         plan_id=str(uuid4()),
-        created_at=datetime.now(),
+        created_at=datetime.now(timezone.utc),
         operations=operations,
         summary={"move": 1, "rename": 1}
     )
@@ -208,7 +208,7 @@ class TestDryRunSimulation:
         
         plan = ExecutionPlan(
             plan_id=str(uuid4()),
-            created_at=datetime.now(),
+            created_at=datetime.now(timezone.utc),
             operations=[no_op_operation],
             summary={"no-op": 1}
         )
@@ -233,7 +233,7 @@ class TestDryRunSimulation:
         
         plan = ExecutionPlan(
             plan_id=str(uuid4()),
-            created_at=datetime.now(),
+            created_at=datetime.now(timezone.utc),
             operations=[conflicted_operation],
             summary={"move": 1}
         )
@@ -285,7 +285,7 @@ class TestExecuteMode:
         
         plan = ExecutionPlan(
             plan_id=str(uuid4()),
-            created_at=datetime.now(),
+            created_at=datetime.now(timezone.utc),
             operations=[operation],
             summary={"move": 1}
         )
@@ -319,7 +319,7 @@ class TestExecuteMode:
         
         plan = ExecutionPlan(
             plan_id=str(uuid4()),
-            created_at=datetime.now(),
+            created_at=datetime.now(timezone.utc),
             operations=[operation],
             summary={"move": 1}
         )
@@ -352,7 +352,7 @@ class TestExecuteMode:
         
         plan = ExecutionPlan(
             plan_id=str(uuid4()),
-            created_at=datetime.now(),
+            created_at=datetime.now(timezone.utc),
             operations=[conflicted_operation],
             summary={"move": 1}
         )
@@ -427,7 +427,7 @@ class TestRollbackLog:
         
         plan = ExecutionPlan(
             plan_id=str(uuid4()),
-            created_at=datetime.now(),
+            created_at=datetime.now(timezone.utc),
             operations=operations,
             summary={"move": 2}
         )
@@ -480,7 +480,7 @@ class TestExecutionSummary:
         
         plan = ExecutionPlan(
             plan_id=str(uuid4()),
-            created_at=datetime.now(),
+            created_at=datetime.now(timezone.utc),
             operations=operations,
             summary={"move": 2, "no-op": 1}
         )
@@ -573,7 +573,7 @@ class TestRollbackLogSaving:
             data = json.load(f)
         
         # Verify timestamps are in ISO format
-        from datetime import datetime
+        from datetime import datetime, timezone
         executed_at = datetime.fromisoformat(data["executed_at"])
         assert executed_at is not None
         
@@ -670,7 +670,7 @@ class TestRollbackExecution:
         
         plan = ExecutionPlan(
             plan_id=str(uuid4()),
-            created_at=datetime.now(),
+            created_at=datetime.now(timezone.utc),
             operations=operations,
             summary={"move": 2}
         )
@@ -759,7 +759,7 @@ class TestRollbackExecution:
         
         plan = ExecutionPlan(
             plan_id=str(uuid4()),
-            created_at=datetime.now(),
+            created_at=datetime.now(timezone.utc),
             operations=operations,
             summary={"move": 1, "no-op": 1}
         )
@@ -877,7 +877,7 @@ class TestRollbackExecution:
         
         plan = ExecutionPlan(
             plan_id=str(uuid4()),
-            created_at=datetime.now(),
+            created_at=datetime.now(timezone.utc),
             operations=operations,
             summary={"move": 2}
         )
@@ -920,7 +920,7 @@ class TestRollbackExecution:
 
         plan = ExecutionPlan(
             plan_id=str(uuid4()),
-            created_at=datetime.now(),
+            created_at=datetime.now(timezone.utc),
             operations=[
                 FileOperation(
                     operation_type="quarantine",

tests/test_planner.py

@@ -1,7 +1,7 @@
 """Unit tests for plan generator."""
 
 import pytest
-from datetime import datetime
+from datetime import datetime, timezone
 from pathlib import Path
 
 from vlm.config import Config
@@ -36,7 +36,7 @@ def test_generate_plan_for_movie_with_year(config):
         path=Path("/mnt/nas/videos/movie/Some.Movie.2020.1080p.mkv"),
         filename="Some.Movie.2020.1080p.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie"
     )
     
@@ -67,7 +67,7 @@ def test_generate_plan_for_movie_without_year(config):
         path=Path("/mnt/nas/videos/movie/random_movie.mkv"),
         filename="random_movie.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie"
     )
     
@@ -94,7 +94,7 @@ def test_generate_plan_for_series_with_season_and_episode(config):
         path=Path("/mnt/nas/videos/series/Show.Name.S01E05.mkv"),
         filename="Show.Name.S01E05.mkv",
         size_bytes=500000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="series"
     )
     
@@ -123,7 +123,7 @@ def test_generate_plan_blocks_series_with_high_season(config):
         path=Path("/mnt/nas/videos/series/Show.Name.S20E01.mkv"),
         filename="Show.Name.S20E01.mkv",
         size_bytes=500000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="series"
     )
     identity = SeriesIdentity(
@@ -147,7 +147,7 @@ def test_generate_plan_blocks_series_with_high_episode(config):
         path=Path("/mnt/nas/videos/series/Show.Name.S01E120.mkv"),
         filename="Show.Name.S01E120.mkv",
         size_bytes=500000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="series"
     )
     identity = SeriesIdentity(
@@ -171,7 +171,7 @@ def test_generate_plan_sample_is_noop_by_default(config):
         path=Path("/mnt/nas/videos/series/Show.Name.Sample.S01E01.mkv"),
         filename="Show.Name.Sample.S01E01.mkv",
         size_bytes=500000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="series"
     )
     identity = SeriesIdentity(
@@ -196,7 +196,7 @@ def test_generate_plan_sample_can_be_included_via_config(config):
         path=Path("/mnt/nas/videos/series/Show.Name.Sample.S01E01.mkv"),
         filename="Show.Name.Sample.S01E01.mkv",
         size_bytes=500000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="series"
     )
     identity = SeriesIdentity(
@@ -219,7 +219,7 @@ def test_generate_plan_for_series_without_season(config):
         path=Path("/mnt/nas/videos/series/ambiguous_show.mkv"),
         filename="ambiguous_show.mkv",
         size_bytes=500000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="series"
     )
     
@@ -247,7 +247,7 @@ def test_generate_plan_for_anime_category(config):
         path=Path("/mnt/nas/videos/anime/Some.Anime.01.mkv"),
         filename="Some.Anime.01.mkv",
         size_bytes=500000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="anime"
     )
     
@@ -267,7 +267,7 @@ def test_generate_plan_for_other_category(config):
         path=Path("/mnt/nas/videos/other/random.mkv"),
         filename="random.mkv",
         size_bytes=500000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="other"
     )
     
@@ -286,7 +286,7 @@ def test_generate_plan_preserves_category_boundaries(config):
         path=Path("/mnt/nas/videos/movie/Movie.2020.mkv"),
         filename="Movie.2020.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie"
     )
     
@@ -311,7 +311,7 @@ def test_generate_plan_for_multi_episode_file(config):
         path=Path("/mnt/nas/videos/series/Show.S01E01E02.mkv"),
         filename="Show.S01E01E02.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="series"
     )
     
@@ -338,7 +338,7 @@ def test_generate_plan_file_already_at_target(config):
         path=Path("/mnt/nas/videos/movie/Some Movie (2020)/Some Movie (2020).mkv"),
         filename="Some Movie (2020).mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie"
     )
     
@@ -364,7 +364,7 @@ def test_generate_plan_rename_vs_move(config):
         path=Path("/mnt/nas/videos/movie/Some Movie (2020)/old_name.mkv"),
         filename="old_name.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie"
     )
     
@@ -385,7 +385,7 @@ def test_generate_plan_rename_vs_move(config):
         path=Path("/mnt/nas/videos/movie/wrong_dir/Some Movie (2020).mkv"),
         filename="Some Movie (2020).mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie"
     )
     
@@ -403,7 +403,7 @@ def test_generate_plan_summary(config):
                 path=Path("/mnt/nas/videos/movie/Movie1.2020.mkv"),
                 filename="Movie1.2020.mkv",
                 size_bytes=1000000,
-                modified_timestamp=datetime.now(),
+                modified_timestamp=datetime.now(timezone.utc),
                 category="movie"
             ),
             MovieIdentity(
@@ -420,7 +420,7 @@ def test_generate_plan_summary(config):
                 path=Path("/mnt/nas/videos/movie/Movie2.mkv"),
                 filename="Movie2.mkv",
                 size_bytes=1000000,
-                modified_timestamp=datetime.now(),
+                modified_timestamp=datetime.now(timezone.utc),
                 category="movie"
             ),
             MovieIdentity(
@@ -437,7 +437,7 @@ def test_generate_plan_summary(config):
                 path=Path("/mnt/nas/videos/anime/Anime.01.mkv"),
                 filename="Anime.01.mkv",
                 size_bytes=500000,
-                modified_timestamp=datetime.now(),
+                modified_timestamp=datetime.now(timezone.utc),
                 category="anime"
             ),
             None
@@ -448,7 +448,7 @@ def test_generate_plan_summary(config):
                 path=Path("/mnt/nas/videos/series/Show.S01E01.mkv"),
                 filename="Show.S01E01.mkv",
                 size_bytes=500000,
-                modified_timestamp=datetime.now(),
+                modified_timestamp=datetime.now(timezone.utc),
                 category="series"
             ),
             SeriesIdentity(
@@ -480,14 +480,14 @@ def test_generate_plan_with_analysis_by_quality(config):
         path=p_720,
         filename="Test.2020.720p.WEB-DL.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie",
     )
     vf_1080 = VideoFile(
         path=p_1080,
         filename="Test.2020.1080p.BluRay.mkv",
         size_bytes=2000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie",
     )
     identity = MovieIdentity(
@@ -528,14 +528,14 @@ def test_generate_plan_with_analysis_by_reputation_quality_time_reason(config):
         path=p_old,
         filename="Test.2020.720p.WEB-DL.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie",
     )
     vf_new = VideoFile(
         path=p_new,
         filename="Test.2020.1080p.BluRay.mkv",
         size_bytes=2000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie",
     )
     identity = MovieIdentity(
@@ -574,14 +574,14 @@ def test_generate_plan_with_analysis_by_reputation_missing_scores_uses_fallback_
         path=p_web,
         filename="Test.2020.2160p.WEB-DL.mkv",
         size_bytes=3000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie",
     )
     vf_bluray = VideoFile(
         path=p_bluray,
         filename="Test.2020.1080p.BluRay.mkv",
         size_bytes=2000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie",
     )
     identity = MovieIdentity(
@@ -622,14 +622,14 @@ def test_generate_plan_human_summary_marks_disc_files_as_high_risk(config):
         path=p_disc1,
         filename="The.Best.of.Youth.DISC1.mkv",
         size_bytes=2000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie",
     )
     vf2 = VideoFile(
         path=p_disc2,
         filename="The.Best.of.Youth.DISC2.mkv",
         size_bytes=1800000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie",
     )
     identity = MovieIdentity(
@@ -669,7 +669,7 @@ def test_generate_plan_with_different_extensions(config):
             path=Path(f"/mnt/nas/videos/movie/Movie.2020{ext}"),
             filename=f"Movie.2020{ext}",
             size_bytes=1000000,
-            modified_timestamp=datetime.now(),
+            modified_timestamp=datetime.now(timezone.utc),
             category="movie"
         )
         
@@ -704,7 +704,7 @@ def test_conflict_detection_for_movie(config, tmp_path):
         path=tmp_path / "movie" / "Some.Movie.2020.1080p.mkv",
         filename="Some.Movie.2020.1080p.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie"
     )
     
@@ -741,7 +741,7 @@ def test_conflict_detection_for_series(config, tmp_path):
         path=tmp_path / "series" / "Show.Name.S01E05.1080p.mkv",
         filename="Show.Name.S01E05.1080p.mkv",
         size_bytes=500000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="series"
     )
     
@@ -776,7 +776,7 @@ def test_no_conflict_when_destination_does_not_exist(config, tmp_path):
         path=source_dir / "Some.Movie.2020.mkv",
         filename="Some.Movie.2020.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie"
     )
     
@@ -814,7 +814,7 @@ def test_conflict_detection_with_multiple_files(config, tmp_path):
                 path=tmp_path / "movie" / "Movie1.2020.mkv",
                 filename="Movie1.2020.mkv",
                 size_bytes=1000000,
-                modified_timestamp=datetime.now(),
+                modified_timestamp=datetime.now(timezone.utc),
                 category="movie"
             ),
             MovieIdentity(
@@ -831,7 +831,7 @@ def test_conflict_detection_with_multiple_files(config, tmp_path):
                 path=tmp_path / "movie" / "Movie2.2021.mkv",
                 filename="Movie2.2021.mkv",
                 size_bytes=1000000,
-                modified_timestamp=datetime.now(),
+                modified_timestamp=datetime.now(timezone.utc),
                 category="movie"
             ),
             MovieIdentity(
@@ -864,7 +864,7 @@ def test_save_plan_to_json(config, tmp_path):
         path=Path("/mnt/nas/videos/movie/Movie.2020.mkv"),
         filename="Movie.2020.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie"
     )
     
@@ -909,7 +909,7 @@ def test_load_plan_from_json(config, tmp_path):
         path=Path("/mnt/nas/videos/movie/Movie.2020.mkv"),
         filename="Movie.2020.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie"
     )
     
@@ -963,7 +963,7 @@ def test_save_and_load_plan_with_conflicts(config, tmp_path):
         path=tmp_path / "movie" / "Movie.2020.mkv",
         filename="Movie.2020.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie"
     )
     
@@ -1003,7 +1003,7 @@ def test_save_and_load_plan_with_no_op_operations(config, tmp_path):
                 path=Path("/mnt/nas/videos/anime/Anime.01.mkv"),
                 filename="Anime.01.mkv",
                 size_bytes=500000,
-                modified_timestamp=datetime.now(),
+                modified_timestamp=datetime.now(timezone.utc),
                 category="anime"
             ),
             None
@@ -1014,7 +1014,7 @@ def test_save_and_load_plan_with_no_op_operations(config, tmp_path):
                 path=Path("/mnt/nas/videos/movie/Movie.mkv"),
                 filename="Movie.mkv",
                 size_bytes=1000000,
-                modified_timestamp=datetime.now(),
+                modified_timestamp=datetime.now(timezone.utc),
                 category="movie"
             ),
             MovieIdentity(
@@ -1048,7 +1048,7 @@ def test_save_plan_json_is_human_readable(config, tmp_path):
         path=Path("/mnt/nas/videos/movie/Movie.2020.mkv"),
         filename="Movie.2020.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie"
     )
     
@@ -1089,7 +1089,7 @@ def test_save_plan_with_multiple_operations(config, tmp_path):
                 path=Path("/mnt/nas/videos/movie/Movie1.2020.mkv"),
                 filename="Movie1.2020.mkv",
                 size_bytes=1000000,
-                modified_timestamp=datetime.now(),
+                modified_timestamp=datetime.now(timezone.utc),
                 category="movie"
             ),
             MovieIdentity(
@@ -1106,7 +1106,7 @@ def test_save_plan_with_multiple_operations(config, tmp_path):
                 path=Path("/mnt/nas/videos/series/Show.S01E01.mkv"),
                 filename="Show.S01E01.mkv",
                 size_bytes=500000,
-                modified_timestamp=datetime.now(),
+                modified_timestamp=datetime.now(timezone.utc),
                 category="series"
             ),
             SeriesIdentity(
@@ -1124,7 +1124,7 @@ def test_save_plan_with_multiple_operations(config, tmp_path):
                 path=Path("/mnt/nas/videos/anime/Anime.01.mkv"),
                 filename="Anime.01.mkv",
                 size_bytes=500000,
-                modified_timestamp=datetime.now(),
+                modified_timestamp=datetime.now(timezone.utc),
                 category="anime"
             ),
             None
@@ -1176,7 +1176,7 @@ def test_plan_json_includes_all_required_fields(config, tmp_path):
         path=Path("/mnt/nas/videos/movie/Movie.2020.mkv"),
         filename="Movie.2020.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie"
     )
     
@@ -1214,7 +1214,7 @@ def test_movie_rejected_by_review_generates_noop(config):
         path=Path("/mnt/nas/videos/movie/Movie.2020.mkv"),
         filename="Movie.2020.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="movie"
     )
 
@@ -1238,7 +1238,7 @@ def test_series_rejected_by_review_generates_noop(config):
         path=Path("/mnt/nas/videos/series/Show.S01E01.mkv"),
         filename="Show.S01E01.mkv",
         size_bytes=1000000,
-        modified_timestamp=datetime.now(),
+        modified_timestamp=datetime.now(timezone.utc),
         category="series"
     )
 

tests/test_quarantine.py

@@ -3,7 +3,7 @@
 import json
 import pytest
 from pathlib import Path
-from datetime import datetime
+from datetime import datetime, timezone
 
 from vlm.quarantine import QuarantineManager
 from vlm.config import Config

tests/test_reports.py

@@ -185,7 +185,7 @@ class TestInventoryReport:
                 Path("/test.mkv"),
                 "test.mkv",
                 1000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie"
             )
         ]
@@ -377,7 +377,7 @@ class TestDuplicateReport:
                 Path("/movies/The.Matrix.1999.1080p.mkv"),
                 "The.Matrix.1999.1080p.mkv",
                 2000000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie",
                 resolution="1920x1080",
                 codec="h264",
@@ -388,7 +388,7 @@ class TestDuplicateReport:
                 Path("/movies/The.Matrix.1999.720p.mkv"),
                 "The.Matrix.1999.720p.mkv",
                 1000000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie",
                 resolution="1280x720",
                 codec="h264"
@@ -448,14 +448,14 @@ class TestDuplicateReport:
                 Path("/movies/Inception.2010.1080p.mkv"),
                 "Inception.2010.1080p.mkv",
                 2000000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie"
             ),
             VideoFile(
                 Path("/movies/Inception.2010.720p.mkv"),
                 "Inception.2010.720p.mkv",
                 1000000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie"
             ),
         ]
@@ -501,14 +501,14 @@ class TestDuplicateReport:
                 Path("/series/Breaking.Bad.S01E01.1080p.mkv"),
                 "Breaking.Bad.S01E01.1080p.mkv",
                 1500000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "series"
             ),
             VideoFile(
                 Path("/series/Breaking.Bad.S01E01.720p.mkv"),
                 "Breaking.Bad.S01E01.720p.mkv",
                 800000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "series"
             ),
         ]
@@ -564,8 +564,8 @@ class TestDuplicateReport:
         # Create two duplicate groups with different sizes
         identity1 = MovieIdentity("Small Movie", 2020, 0.9, False, "Small.Movie.mkv")
         files1 = [
-            VideoFile(Path("/movies/Small.Movie.1.mkv"), "Small.Movie.1.mkv", 500000000, datetime.now(), "movie"),
-            VideoFile(Path("/movies/Small.Movie.2.mkv"), "Small.Movie.2.mkv", 600000000, datetime.now(), "movie"),
+            VideoFile(Path("/movies/Small.Movie.1.mkv"), "Small.Movie.1.mkv", 500000000, datetime.now(timezone.utc), "movie"),
+            VideoFile(Path("/movies/Small.Movie.2.mkv"), "Small.Movie.2.mkv", 600000000, datetime.now(timezone.utc), "movie"),
         ]
         quality1 = [
             {'filename': 'Small.Movie.1.mkv', 'path': '/movies/Small.Movie.1.mkv', 'size_bytes': 500000000},
@@ -574,8 +574,8 @@ class TestDuplicateReport:
         
         identity2 = MovieIdentity("Large Movie", 2021, 0.9, False, "Large.Movie.mkv")
         files2 = [
-            VideoFile(Path("/movies/Large.Movie.1.mkv"), "Large.Movie.1.mkv", 2000000000, datetime.now(), "movie"),
-            VideoFile(Path("/movies/Large.Movie.2.mkv"), "Large.Movie.2.mkv", 1800000000, datetime.now(), "movie"),
+            VideoFile(Path("/movies/Large.Movie.1.mkv"), "Large.Movie.1.mkv", 2000000000, datetime.now(timezone.utc), "movie"),
+            VideoFile(Path("/movies/Large.Movie.2.mkv"), "Large.Movie.2.mkv", 1800000000, datetime.now(timezone.utc), "movie"),
         ]
         quality2 = [
             {'filename': 'Large.Movie.1.mkv', 'path': '/movies/Large.Movie.1.mkv', 'size_bytes': 2000000000},
@@ -600,8 +600,8 @@ class TestDuplicateReport:
         """Sort order should use quality_comparison sizes when VideoFile sizes are zero."""
         identity1 = MovieIdentity("Tiny", 2020, 0.9, False, "Tiny.mkv")
         files1 = [
-            VideoFile(Path("/movies/Tiny.1.mkv"), "Tiny.1.mkv", 0, datetime.now(), "movie"),
-            VideoFile(Path("/movies/Tiny.2.mkv"), "Tiny.2.mkv", 0, datetime.now(), "movie"),
+            VideoFile(Path("/movies/Tiny.1.mkv"), "Tiny.1.mkv", 0, datetime.now(timezone.utc), "movie"),
+            VideoFile(Path("/movies/Tiny.2.mkv"), "Tiny.2.mkv", 0, datetime.now(timezone.utc), "movie"),
         ]
         quality1 = [
             {"filename": "Tiny.1.mkv", "path": "/movies/Tiny.1.mkv", "size_bytes": 600000000},
@@ -610,8 +610,8 @@ class TestDuplicateReport:
 
         identity2 = MovieIdentity("Huge", 2021, 0.9, False, "Huge.mkv")
         files2 = [
-            VideoFile(Path("/movies/Huge.1.mkv"), "Huge.1.mkv", 0, datetime.now(), "movie"),
-            VideoFile(Path("/movies/Huge.2.mkv"), "Huge.2.mkv", 0, datetime.now(), "movie"),
+            VideoFile(Path("/movies/Huge.1.mkv"), "Huge.1.mkv", 0, datetime.now(timezone.utc), "movie"),
+            VideoFile(Path("/movies/Huge.2.mkv"), "Huge.2.mkv", 0, datetime.now(timezone.utc), "movie"),
         ]
         quality2 = [
             {"filename": "Huge.1.mkv", "path": "/movies/Huge.1.mkv", "size_bytes": 3000000000},
@@ -632,12 +632,12 @@ class TestSummaryReport:
     def test_generate_summary_report(self):
         """Test generating summary report with various files."""
         files = [
-            VideoFile(Path("/movies/Movie1.mkv"), "Movie1.mkv", 2000000000, datetime.now(), "movie"),
-            VideoFile(Path("/movies/Movie2.mkv"), "Movie2.mkv", 1500000000, datetime.now(), "movie"),
-            VideoFile(Path("/series/Show.S01E01.mkv"), "Show.S01E01.mkv", 1000000000, datetime.now(), "series"),
-            VideoFile(Path("/series/Show.S01E02.mkv"), "Show.S01E02.mkv", 1000000000, datetime.now(), "series"),
-            VideoFile(Path("/anime/Anime1.mkv"), "Anime1.mkv", 800000000, datetime.now(), "anime"),
-            VideoFile(Path("/other/Random.mkv"), "Random.mkv", 500000000, datetime.now(), "other"),
+            VideoFile(Path("/movies/Movie1.mkv"), "Movie1.mkv", 2000000000, datetime.now(timezone.utc), "movie"),
+            VideoFile(Path("/movies/Movie2.mkv"), "Movie2.mkv", 1500000000, datetime.now(timezone.utc), "movie"),
+            VideoFile(Path("/series/Show.S01E01.mkv"), "Show.S01E01.mkv", 1000000000, datetime.now(timezone.utc), "series"),
+            VideoFile(Path("/series/Show.S01E02.mkv"), "Show.S01E02.mkv", 1000000000, datetime.now(timezone.utc), "series"),
+            VideoFile(Path("/anime/Anime1.mkv"), "Anime1.mkv", 800000000, datetime.now(timezone.utc), "anime"),
+            VideoFile(Path("/other/Random.mkv"), "Random.mkv", 500000000, datetime.now(timezone.utc), "other"),
         ]
         
         library_root = Path("/mnt/nas/videos")
@@ -677,8 +677,8 @@ class TestSummaryReport:
     def test_generate_summary_report_single_category(self):
         """Test generating summary report with files in single category."""
         files = [
-            VideoFile(Path("/movies/Movie1.mkv"), "Movie1.mkv", 1000000000, datetime.now(), "movie"),
-            VideoFile(Path("/movies/Movie2.mkv"), "Movie2.mkv", 2000000000, datetime.now(), "movie"),
+            VideoFile(Path("/movies/Movie1.mkv"), "Movie1.mkv", 1000000000, datetime.now(timezone.utc), "movie"),
+            VideoFile(Path("/movies/Movie2.mkv"), "Movie2.mkv", 2000000000, datetime.now(timezone.utc), "movie"),
         ]
         
         library_root = Path("/mnt/nas/videos")

tests/test_reports_integration.py

@@ -5,7 +5,7 @@ Tests the complete workflow from analysis to report generation.
 
 import json
 from pathlib import Path
-from datetime import datetime
+from datetime import datetime, timezone
 from vlm.models import SeriesIdentity, VideoFile, MovieIdentity
 from vlm.analysis import analyze_series_completeness, detect_duplicates
 from vlm.reports import generate_completeness_report, generate_duplicate_report, generate_summary_report
@@ -59,7 +59,7 @@ class TestReportsIntegration:
                 Path("/movies/The.Matrix.1999.1080p.mkv"),
                 "The.Matrix.1999.1080p.mkv",
                 2000000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie",
                 resolution="1920x1080",
                 codec="h264"
@@ -68,7 +68,7 @@ class TestReportsIntegration:
                 Path("/movies/The.Matrix.1999.720p.mkv"),
                 "The.Matrix.1999.720p.mkv",
                 1000000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie",
                 resolution="1280x720",
                 codec="h264"
@@ -77,7 +77,7 @@ class TestReportsIntegration:
                 Path("/movies/Inception.2010.mkv"),
                 "Inception.2010.mkv",
                 1500000000,
-                datetime.now(),
+                datetime.now(timezone.utc),
                 "movie"
             ),
         ]
@@ -107,11 +107,11 @@ class TestReportsIntegration:
         """Test summary report generation with mixed file types."""
         # Create test files
         files = [
-            VideoFile(Path("/movies/Movie1.mkv"), "Movie1.mkv", 2000000000, datetime.now(), "movie"),
-            VideoFile(Path("/movies/Movie2.mkv"), "Movie2.mkv", 1500000000, datetime.now(), "movie"),
-            VideoFile(Path("/series/Show.S01E01.mkv"), "Show.S01E01.mkv", 1000000000, datetime.now(), "series"),
-            VideoFile(Path("/series/Show.S01E02.mkv"), "Show.S01E02.mkv", 1000000000, datetime.now(), "series"),
-            VideoFile(Path("/anime/Anime1.mkv"), "Anime1.mkv", 800000000, datetime.now(), "anime"),
+            VideoFile(Path("/movies/Movie1.mkv"), "Movie1.mkv", 2000000000, datetime.now(timezone.utc), "movie"),
+            VideoFile(Path("/movies/Movie2.mkv"), "Movie2.mkv", 1500000000, datetime.now(timezone.utc), "movie"),
+            VideoFile(Path("/series/Show.S01E01.mkv"), "Show.S01E01.mkv", 1000000000, datetime.now(timezone.utc), "series"),
+            VideoFile(Path("/series/Show.S01E02.mkv"), "Show.S01E02.mkv", 1000000000, datetime.now(timezone.utc), "series"),
+            VideoFile(Path("/anime/Anime1.mkv"), "Anime1.mkv", 800000000, datetime.now(timezone.utc), "anime"),
         ]
         
         # Generate summary report

tests/test_state.py

@@ -92,7 +92,7 @@ class TestLoadSaveState:
         store = StateStore(
             states={},
             version='1.0',
-            last_updated=datetime.now()
+            last_updated=datetime.now(timezone.utc)
         )
         
         save_state(store, state_path)