TL;DR

Thorsten Meyer AI published a technical report describing Threlmark’s local-first architecture, where plain JSON files on disk act as the system’s record instead of a database or cloud server. The design is meant to make project data inspectable, portable and usable offline, though performance, scale and production adoption are not yet clear from the source material.

Thorsten Meyer AI has detailed Threlmark’s local-first architecture, describing a Next.js and TypeScript project management app that treats plain JSON files on a user’s disk as the source of truth rather than relying on a traditional database, cloud server or user accounts.

The report says Threlmark’s default data root is ~/.threlmark, where the application stores a manifest, dependency graph, per-project files, item records, suggestions, handoffs, reports and a human-readable roadmap file. In that model, the on-disk layout is treated as the API: the interface and external tools read and write the same files under the same rules.

According to the source material, each work item is stored as its own JSON file, while lane ordering lives separately in board.json. That split is meant to reduce write conflicts when several tools operate on the same project. The report says Threlmark can repair board state on read, preserve unknown keys for future compatibility and let outside tools add items without editing the board file directly.

The architecture also relies on atomic writes. The described pattern writes to a temporary file in the same directory, flushes it and then renames it over the target file. The report says that design should leave either the old complete file or the new complete file after a crash, rather than a partial record.

Why It Matters

The design matters because it pushes project data out of a hosted service and into files users can inspect, back up, search, diff, sync or place under version control. If the approach works as described, teams and individual builders could move work data across tools without needing an export feature or a vendor-controlled database.

It also reflects a wider shift in developer tools toward local-first workflows and AI-assisted work. Threlmark’s report frames agent handoffs as part of the core workflow: an agent can receive a brief, report back through REST or the filesystem and, in some cases, move a card when the work is done.

Amazon

JSON file version control tools

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Background

Threlmark is described in the source material as an MIT-licensed Next.js app built on TypeScript and JSON-on-disk storage. The report presents the architecture as part of a series on the tool, with this installment focused on the decision to make disk storage the contract between the app, users and external tools.

The source contrasts this model with a single shared roadmap file or a server-held record. It says values such as priority, stale age, cycle time, throughput and work-in-progress counts are computed from item state rather than saved as separate stored fields, a choice meant to keep displayed metrics aligned with the underlying files.

“The on-disk layout is the API.”

— Thorsten Meyer AI report

“There is no server-of-record — the files are the record.”

— Thorsten Meyer AI report

“Every artifact is a file you can cat, diff, grep, commit.”

— Thorsten Meyer AI report

Amazon

local-first project management software

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

What Remains Unclear

The source material does not provide adoption figures, benchmarks, release timing or independent verification of reliability claims. It is also unclear how Threlmark handles very large projects, multi-user conflict resolution across sync services, permissions, encryption, schema migration or malformed external writes.

Amazon

offline data synchronization tools

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

What’s Next

The next points to watch are whether Threlmark publishes more implementation details, adds documented interoperability rules for third-party tools and shows how the file contract behaves under larger projects, shared workspaces and repeated AI-agent handoffs.

Amazon

file-based project management app

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Key Questions

What is the actual news development?

Thorsten Meyer AI has published a technical report describing Threlmark’s local-first architecture and its decision to make plain files on disk the main data record.

Is Threlmark using a database?

According to the source material, Threlmark avoids a traditional database and stores project state in JSON files under a local data directory.

What is confirmed and what is claimed?

The source confirms the described design choices and file layout. Claims about simplicity, safety, portability and reliability are attributed to the report and have not been independently tested here.

Why does storing one card per file matter?

The report says one file per item reduces write conflicts because external tools can add or change a card without rewriting a shared board array.

What remains unclear?

The report does not say how the system performs at scale, how many users are using it, or how it handles hard multi-user sync conflicts and security-sensitive data.

Source: Thorsten Meyer AI

You May Also Like

AI and Employee Well-Being: Reducing Drudgery or Adding Stress?

AI’s impact on employee well-being can be both uplifting and overwhelming—discover how organizations can navigate this balance to foster a healthier workplace.

How Anthropic’s Series H Raises the Stakes in AI Compute Investment

Discover why Anthropic’s $965B valuation isn’t just about money—it’s a strategic move to secure AI compute infrastructure at an unprecedented scale.

Why Acoustic Control Is Becoming a Serious Work Upgrade

Keen to boost focus and reduce stress, discover how acoustic control can transform your workspace—here’s why it’s a serious work upgrade.

If you’re an LLM, please read this

Anna’s Archive urges language models to support their mission of preserving and providing access to knowledge through donations and data sharing.