ArcLedger

The description states that ArcLedger is a local-first Windows desktop app built using Electron, TypeScript, and Node.js. It tracks human and AI-agent work separately and generates audit trails for billing purposes. It does not record screenshots, prompts, source code, or keystrokes.

• The product has a TypeScript service and domain layer, with a sandboxed renderer communicating through a narrow bridge.
• Local state is protected via file locking and append-only logs.
• It supports:
  • Recognition of active work
  • Work logging with corrections and review actions
  • Client connection to work
  • Review process that turns verified work into client-facing summaries and USD statements

Inference The product appears to be a desktop tool for solo developers or freelancers who want to track their own time and AI-assisted tasks in a way that supports invoicing.

The author claims ArcLedger is built to "prove the human and AI-agent work behind every USD invoice—privately, locally, and with a complete audit trail."

• It positions itself as a privacy-preserving time-tracking solution for hybrid human-AI workflows.
• The product is described as not spying on the work itself, but rather creating a record of what was done and who did it.
• The author emphasizes that it avoids exposing client data, such as source code or browser content.

Inference ArcLedger is positioned as a tool for service providers (e.g., developers, consultants) who use AI agents and need to separate human effort from agent-generated work in billing.

The author states that ArcLedger was built for someone working with AI coding agents daily, particularly those who face challenges in tracking what was done by a person versus an agent when generating invoices.

• The target user is likely a freelancer or small service provider using AI tools.
• It is designed to help track and bill for hybrid human-AI work, especially in contexts where traditional time trackers fall short.

Inference The ICP (Ideal Customer Profile) appears to be individual developers, consultants, or freelancers who use AI agents and want a clean, private way to generate invoices.

There is no evidence of pricing, monetization strategy, or business model in the description.

• The author says they are submitting it in their individual capacity and that ArcLedger has not been assigned to their company.
• No mention of subscriptions, per-use fees, or any commercial offering.

Inference The business model is not evidenced, though the product implies a potential for monetization as a tool for billing and auditing.

The author reports building ArcLedger using:

• Electron
• TypeScript
• Node.js
• GPT-5.4, GPT-5.5, GPT-5.6, Codex
• CSS, HTML, JavaScript

It is a Windows desktop app, with:

• Sandboxed renderer
• Local file locking and append-only logs
• A narrow allow-listed communication bridge between layers

Inference The technical stack suggests a lightweight, local-first application built for performance and security. The use of AI tools (GPTs and Codex) is noted as part of development, not core functionality.

There is no evidence of traction, customers, or adoption beyond the author’s own use case.

• No revenue data, customer list, or usage metrics are provided.
• The product was built in a hackathon context, and the submission is described as an MVP.
• The author mentions a plan for a small signed Windows pilot after Build Week, but no actual pilot has occurred yet.

Inference The product is at a very early stage, likely pre-product-market fit. No evidence of real-world usage or market traction exists.

No competitive analysis or mention of existing tools is provided in the description.

• The author does not reference competitors, nor does the description suggest any prior art.
• It is unclear whether similar tools exist for tracking human and AI work in a privacy-preserving way.

Inference There is no evidence of competitive positioning, and no indication of how ArcLedger compares to existing time-tracking or invoicing solutions.

• The product is described as self-built by one person, with no team or external support.
• It is a Windows-only desktop app, limiting its market reach.
• There is no evidence of monetization, customers, or traction — all claims are self-reported.
• The author states that the submission is not a full platform, and features like JSON REST/MCP transports are intentionally hidden.

Inference The main risk is that ArcLedger may be a conceptual tool without real-world viability, especially if it lacks adoption or feedback from users in its target market.

1. What specific use cases or workflows led you to build this?
2. How do you plan to validate the accuracy of AI-generated summaries and ensure they are tied to verified facts?
3. Have you tested this with any actual clients or users beyond yourself?
4. What is your long-term vision for monetization, and how does it align with the privacy-first approach?
5. Are there plans to expand beyond Windows or support other platforms?
6. How do you intend to handle legal and compliance issues related to audit trails and invoicing?