Chorsu

The description states that Chorsu is a local-first macOS task router. It uses a configurable global shortcut to open a lightweight capture panel over whatever application the user is in. Users type one task in plain English and press Return.

Key features include:

• Natural time phrases like “at 8 PM” or “tomorrow morning” are parsed into local reminders.
• Tasks are ranked into projects using deterministic rules and on-device signals.
• If routing is uncertain, the panel expands into a keyboard-first project chooser.
• Dependency phrases such as “after,” “once,” and “when” can link prerequisites.
• Blocked tasks stay out of the Today view until their prerequisite is complete, then resurface with local notifications.
• The main interface focuses on Today, Upcoming, and Projects views.
• Tasks can be completed via notifications, menu-bar list, or keyboard shortcuts.

The app stores everything locally on the Mac and does not require an account, cloud service, or API key.

Evidence All of this is self-reported by the author. No third-party verification or usage data is provided.

The description states that Chorsu aims to "protect the thought they are supposed to protect", contrasting with most task apps that interrupt the user’s workflow. It positions itself as a tool for capturing tasks quickly without needing to choose projects, schedule, or remember follow-ups.

It also claims:

• No account, no cloud, no API keys.
• One capture creates exactly one task — no inbox to clean up later.
• Uses deterministic parsing and local signals before resorting to AI.
• On-device models assist asynchronously when available but do not block the capture flow.
• The app is designed for keyboard-first workflows, with support for menu bar visibility and notifications.

This suggests a shift from traditional task managers toward a more integrated, frictionless, and privacy-focused approach.

Evidence These are claims made by the author. No evidence of prior product adoption or customer feedback is provided.

The description does not name specific customers or personas. However, it implies a target audience that:

• Works on macOS.
• Values local-first, privacy-preserving tools.
• Prefers keyboard-first workflows.
• Needs to capture tasks quickly without interrupting their current activity.
• May be interested in dependency-aware task management.

It also suggests users who are comfortable with natural language input and want automation that doesn’t silently misfile work.

Evidence No explicit customer segmentation or persona data. The positioning is inferred from the stated features and use case.

There is no evidence of a business model or pricing structure in the description. The app is described as:

• Free to use.
• Local-only with no cloud, API keys, or accounts.
• Built for personal productivity, not enterprise or commercial use.

The author does not mention monetization plans, subscriptions, or paid features.

Evidence Not evidenced. This is a self-reported product with no indication of revenue streams.

The app is built natively in Swift 6, using:

• SwiftUI for the main interface.
• AppKit for the fast floating capture panel.
• SwiftData for persistence.
• UserNotifications for reminders and alerts.

It uses:

• Deterministic parsing and local project ranking.
• Apple Foundation Models for optional assistance (asynchronous, untrusted input).
• Atomic transactions for task creation and updates.
• Additive schema changes and immutable snapshots.
• Extensive Swift Testing coverage.
• Offline usability when on-device intelligence is unavailable.

The author reports using Codex as an engineering collaborator during Build Week to refactor and implement features like semantic project ranking, temporal parsing, and concurrency hardening.

Evidence All technical details are self-reported. No evidence of production deployment or performance metrics.

There is no evidence of traction, adoption, or user base. The project:

• Was submitted to a hackathon (Devpost).
• Has only one team member.
• Is described as a prototype built during Build Week.
• Has no mention of downloads, users, or feedback.

The author notes that the app was preserved from a pre-Build Week baseline, but there is no indication it has been used beyond development or testing.

Evidence Not evidenced. No data on usage, adoption, or product maturity beyond prototype stage.

The description does not reference any competitors directly. However, based on the stated features — local-first task management, keyboard-first capture, dependency-aware workflows, and no cloud requirements — Chorsu appears to be positioned in a niche that overlaps with:

• Local-first productivity tools.
• Minimalist task managers (e.g., Things, Todoist, Notion).
• AI-assisted task routing (e.g., Obsidian, Roam Research, AI-powered task apps).

It differs from many competitors by emphasizing:

• No account or cloud.
• Deterministic parsing over AI-driven decisions.
• Keyboard-first and notification-based workflows.

Evidence No competitive analysis or market positioning data. This is inferred from the stated features.

Key risks and red flags based on the description:

1. No traction or user feedback: The app appears to be a prototype, with no evidence of real-world usage.
2. Single founder: The team size is listed as one person, which may limit scalability or product development speed.
3. Limited commercialization: No mention of monetization, pricing, or business model.
4. Unproven market fit: The positioning is novel but untested in the market.
5. Dependency on Apple Intelligence: While optional, reliance on on-device AI models may limit functionality if those are not available or perform poorly.

Evidence All inferences based on self-reported description and lack of external data.

1. What is the current status of Chorsu? Is it a working prototype, or has it been released to users?
2. Have you received any feedback from early adopters or beta testers?
3. Are there plans to monetize the product, and if so, how?
4. How do you plan to scale beyond a single developer?
5. What is your roadmap for expanding functionality or addressing dependencies on Apple Intelligence?
6. Are there any technical limitations or edge cases that have not been addressed in the current version?