put·here

Product Introduction

1. Definition: put·here is a local-first, spatial note-taking and knowledge management application for macOS. Technically, it is a spatial canvas interface layered over the native macOS file system (Finder), functioning as a visual workbench for unstructured information.
2. Core Value Proposition: It exists to solve the problem of premature organization in digital note-taking. Unlike traditional note apps that demand immediate categorization (title, folder, tag), put·here allows users to capture ideas, links, images, and files in a free-form canvas, using spatial positioning as a memory aid. Its core value is enabling frictionless collection and visual connection of thoughts before they are fully formed, with optional AI-assisted research that respects user sovereignty over data.

Main Features

1. Spatial Canvas for Unstructured Capture: The primary interface is an infinite, zoomable canvas. Users can place text notes, web links (saved as .webloc files), images, PDFs, and any other file type via drag-and-drop, paste, or a global keyboard shortcut (⌃⌥P). This spatial organization allows for at-a-glance review and leverages spatial memory, a key differentiator from linear note lists.
2. Local-First, File-Based Architecture: Every put·here canvas is a direct representation of a standard macOS Finder folder. Notes are stored as plain Markdown (.md) files, images retain their original formats, and links are .webloc files. This ensures complete data portability, vendor lock-in avoidance, and the ability to access and edit content with any other application. File system changes (e.g., adding a file via Finder) are reflected in the canvas.
3. Integrated AI Research with Source Provenance: Users can right-click a note (e.g., a question) and "Hand it to AI." The feature creates a separate workspace where the AI reads the provided context, conducts web research, and returns findings as new cards, each citing its source URL. This keeps AI-generated content isolated from the user's main canvas, maintaining user control. It supports the built-in (allowance-based) AI, user-provided API keys (OpenAI, Anthropic, etc.), or local models (LM Studio, Ollama).
4. Visual Grouping and Connection Tools: Users can manually group related cards and draw explicit connection lines between them, with optional labels. A "Tidy" (Beta) feature can algorithmically organize a selected cluster of cards. All spatial arrangements and connections are saved as metadata, allowing for non-destructive reorganization.

Problems Solved

1. Pain Point: The cognitive friction and loss of context caused by forcing immediate structure onto nascent ideas. The "scattered research" problem, where related information is trapped across dozens of browser tabs, disparate notes, and downloaded files with no visual relationship map.
2. Target Audience: Knowledge Workers & Researchers (academics, journalists, product managers) conducting open-ended research. Students learning complex subjects who need to connect concepts visually. Creative Professionals (writers, designers) in the early, ideation phase of projects. Technically-minded users who prioritize data ownership and plain-text formats.
3. Use Cases: Exploring a product direction by gathering market research, competitor links, and feature ideas on one canvas. Learning a difficult concept like linear algebra by placing textbook PDFs, lecture notes, practice questions, and explanatory links in a relational map. Investigating an open question by collecting all disparate evidence and using AI to research gaps without leaving the context.

Unique Advantages

1. Strengths & Limitations (Pros & Cons):

   • Pros: Unmatched data ownership and portability due to its file-based, local-first model. Exceptionally low-friction capture reduces barrier to saving thoughts. Spatial interface uniquely supports non-linear, associative thinking. AI integration is thoughtfully designed to be assistive, not autonomous, with strong source citation.
   • Cons: Currently macOS-only (14+), excluding Windows and Linux users. The spatial paradigm, while powerful, may not suit users who prefer strictly hierarchical, list-based organization. Advanced publishing/sharing features are limited or gated. Being a relatively new tool, it lacks the extensive plugin ecosystems of established note-taking apps.

2. Key Alternatives & Differentiation:

   • vs. Obsidian: Both are local-first and Markdown-based. Obsidian is a powerful, graph-based editor with a vast plugin ecosystem but a primarily document-centric view. put·here differentiates with its primary spatial canvas interface, treating position as a first-class citizen, and its built-in, canvas-aware AI research workspace.
   • vs. Notion / Evernote: These are cloud-based, database-oriented platforms focused on structured, collaborative workspaces. put·here is the antithesis for personal, pre-structured thought: it offers no databases, requires no account, stores files locally, and emphasizes free-form spatial placement over predefined templates and properties.
   • vs. Miro / FigJam: These are real-time collaborative whiteboards. put·here is for personal knowledge management, with deeper file system integration (real folders/Markdown) and AI research features, but lacks real-time multi-user collaboration.

Frequently Asked Questions (FAQ)

1. Is put·here free, and what are the costs for AI? The core put·here application (v0.13.2) is free to download with no account required. Its built-in AI service includes a monthly free allowance. For extended use, you can supply your own paid API key (costs depend on the provider, e.g., OpenAI) or configure a free local model, giving you full control over AI usage costs and privacy.
2. Where is my put·here data stored, and can I access it without the app? All data is stored in standard folders on your Mac's file system. You can navigate to, open, edit, and back up your notes (Markdown), images, and files directly in macOS Finder or any other application. This guarantees your notes are never locked into the put·here ecosystem.
3. Can I use put·here on iPad, iPhone, or Windows? No, put·here is currently a native application exclusively for macOS (version 14 or later). It runs on both Intel and Apple silicon Macs. There is no web version, iOS app, or Windows client at this time.
4. How does put·here's AI ensure my private data isn't leaked? When using the built-in AI, only the text necessary for the specific research query is sent to put·here's server, linked to an anonymous device ID for allowance counting. If you use your own API key, requests go directly from your Mac to your chosen provider (e.g., OpenAI). If you use a local model, all processing happens entirely on your device. AI always works in a separate workspace, not your main canvas.
5. How do I share a canvas I've created in put·here? The primary method is exporting a "share bundle." The hosted publishing feature to generate a shareable web link is currently in a limited beta and requires an invitation code; it is not part of the standard download experience. For collaboration, you can place the canvas folder in a synced service (iCloud Drive, Dropbox) for others to open in their own put·here instance.