Multimedia Manager (Community Edition)
Version 1.0.6 | Author: nuoguiAI | Changelog
A self-hosted multimedia management agent — organize, search, and browse all your images and videos with a beautiful web gallery. All data stays local on your machine.
多媒体管理 Agent 社区版 — 图片视频管理、搜索、Web 画廊。数据完全本地存储。
Features
- Import: Images and videos with auto-deduplication (SHA256)
- Search: Full-text search (FTS5) across descriptions and tags
- Gallery: Responsive web gallery with infinite scroll
- Albums: Create and manage photo albums
- Favorites: Star your best photos
- CLI: Command-line tools for import, search, and management
- Categories: Organize by type (ai-generated, family, screenshots, etc.)
Pro Edition (Available Separately)
Contact: xujk2012@gmail.com
Upgrade for advanced features:
- AI auto-tagging (Qwen-VL vision model)
- Face recognition and clustering (InsightFace)
- GPS map view with photo markers
- Semantic search with embeddings
- Timeline and Memories
- Duplicate detection
- Resumable chunked uploads
- Feishu/Lark integration
- Share links and multi-user support
Security & Permissions
This skill runs entirely on your local machine. Here's what it accesses and why:
| Resource | Why | Scope |
|---|---|---|
| File system | Read/write images to ~/.image-vault | Only IMAGE_VAULT_DIR directory |
| SQLite database | Store image metadata (no images in DB) | ~/.image-vault/image_vault.db |
| Local HTTP server | Web gallery UI on 127.0.0.1 | Localhost only, token-protected |
| Python packages | Flask, Pillow, PyYAML (via pip) | Standard packages |
No data leaves your machine after initial setup. No telemetry, no cloud sync, no external API calls. The setup.sh script performs a one-time pip install (requires network) to install 3 Python packages (Flask, Pillow, PyYAML) and creates the vault directory. You can review it before running or install dependencies manually.
Environment Variables
All environment variables used by this skill are explicitly declared below. See .env.example for a ready-to-copy template.
| Variable | Required | Description | Default |
|---|---|---|---|
IMAGE_VAULT_TOKEN | Yes | Authentication token for the web gallery. Auto-generated by setup.sh. | _(none)_ |
IMAGE_VAULT_DIR | No | Root directory for image storage. | ~/.image-vault |
IMAGE_VAULT_HOST | No | Server bind address. | 127.0.0.1 |
IMAGE_VAULT_DB | No | Path to the SQLite database file. | <IMAGE_VAULT_DIR>/image_vault.db |
PORT | No | Web gallery server port. | 9876 |
These are the only environment variables read by this skill.
Behavior Transparency
The following behaviors may be flagged by security scanners. All are intentional and required for the skill to function:
| Behavior | Why | Details |
|---|---|---|
setup.sh runs pip install | One-time dependency install | Installs Flask, Pillow, PyYAML from PyPI. Requires network. You can install manually instead: pip install Flask Pillow PyYAML |
Reads .env file | Load config without exporting shell vars | Standard dotenv pattern. Only reads from the skill's own directory. See .env.example for all supported keys |
cleanup_inbound.py deletes files | Remove processed import files | Only deletes files in <IMAGE_VAULT_DIR>/inbound/ (configurable via --dir). Controlled via --days flag (default: 7 days) |
Subprocess calls (ffmpeg/ffprobe) | Extract video metadata and thumbnails | Optional — video features degrade gracefully if these binaries are not installed. No other subprocess calls are made |
Quick Start
# 1. Review and run setup (installs Flask, Pillow, PyYAML)
bash setup.sh
# 2. Start gallery (localhost only by default)
python3 scripts/image_cli.py serve --port 9876Open http://localhost:9876 and enter your token.
CLI Commands
# Import
python3 scripts/image_cli.py import /path/to/photos --source uploaded
# Search
python3 scripts/image_cli.py search "sunset" --json
# Stats
python3 scripts/image_cli.py stats
# Start gallery
python3 scripts/image_cli.py serve --port 9876