makecindy
Featured
cindy
Cindy is an open-source AI agent that runs locally on your machine, integrates multiple AI harnesses and models, and provides memory, skills, and automation to perform real work in your projects and apps.
AI repository intelligence
FR-AI / ANALYSISWhy it is worth attention
It combines predictable format‑based sorting with optional AI‑powered content classification, all running entirely offline with no telemetry or accounts.
Who it is for
- macOS users who want automated Downloads cleanup
- Privacy‑conscious individuals avoiding cloud AI
- Users with mixed document types needing subject‑based sorting
- Developers interested in on‑device Foundation Model integration
Use cases
- Automatically sort downloaded PDFs, images, and code by file type
- Classify university, finance, or medical documents by content without uploading
- Undo accidental moves and review uncertain files before organization
- Detect exact duplicate files in Downloads with SHA‑256 and move extras to Trash
Strengths
- Local‑first: no accounts, analytics, or network calls for classification
- Two‑mode system (Format mode for reliability, Smart Content for subject folders)
- Full undo support and collision‑safe file moves
- Configurable categories, teachable on‑device model, and import/export profiles
Considerations
- Requires macOS 26 and Apple silicon for Smart Content mode
- Smart Content model can make incorrect suggestions and needs user review
- Not notarized; users must bypass Gatekeeper for GitHub releases
README quick start
FileMorrow
Organize your Downloads with on-device Apple Intelligence—guided by your own rules.
Private, reversible, and made for macOS by Nabeegh.
FileMorrow keeps fresh files in Today, Yesterday, and Last 7 Days views. Only older loose files become eligible for category folders. Downloaded folders are never moved or reorganized, and every organization batch can be undone.
| Format mode | Smart Content mode | |
|---|---|---|
| Default | Yes | Opt-in |
| Classification | Known extension and system file type | Filename, local extracted evidence, then Apple Foundation Models when needed |
| Apple Intelligence required | No | Yes, for unresolved content |
| Predictability | Highest | More meaningful subject folders, but suggestions can be wrong |
| Privacy | Local | Local, using Apple's on-device model |
| Best for | Automatic low-maintenance cleanup | Mixed-subject University, Finance, Medical, Legal, and Work documents |
Format mode is the recommended default. Smart Content is useful when subject matters more than extension, but the model is not perfectly accurate. FileMorrow therefore uses deterministic evidence first and leaves uncertain items visible for review.
Highlights
- On-device classification with Apple Foundation Models
- Predictable By File Format mode with no AI queue
- Optional Smart Content mode using metadata, extracted content, and Apple Intelligence
- Fast local evidence scoring before invoking the language model
- Content extraction for PDF, DOCX, PPTX, XLSX, ZIP, text, code, and images
- Local OCR with Vision
- Persistent AI decisions and user corrections
- Quick Look file previews and one-click opening in the file’s default app
- A Teach Organizer workflow for corrections, filename rules, format rules, and reusable on-device model examples
- Separate Awaiting Analysis and Needs Review states
- Analyze-next and continuous Analyze All modes with a Stop control
- Editable, importable, and exportable organization profiles
- User-selectable and fully custom categories
- Confidence-gated organization
- Plan-first approval before any eligible file moves
- Explicit Undo Last Organization, Command-Z, and collision-safe moves
- Progress-aware, cancellable exact duplicate detection with SHA-256 and recoverable Trash cleanup
- Top-level-only organiz
Description
A private, local-first Downloads organizer for macOS with safe automation, Undo, duplicate cleanup, and optional Apple Intelligence.
Related repositories
Similar projects matched by category, topics, and programming language.
Blaizzy
Featured
nativ
Nativ is a native macOS app that lets you run AI models locally on Apple silicon, offering chat, model management, performance analytics, and an OpenAI/Anthropic-compatible API server.
simonw
Featured
speech-analyzer-cli
A macOS command-line tool that uses Apple's on-device speech APIs to transcribe prerecorded audio into plain text, JSON, JSONL, SRT, or WebVTT.