📊 Full opportunity report: Disk Is the Contract: Inside Threlmark’s Local-First Architecture on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
Threlmark introduces a local-first, disk-based architecture for project management that eliminates the need for a central server. Its design prioritizes data portability, safety, and seamless AI integration, challenging traditional cloud-based tools.
Threlmark has unveiled a novel, local-first architecture that relies entirely on files stored on disk, with no server or cloud dependency, to manage project data and AI workflows. This approach emphasizes data portability, safety, and interoperability, marking a significant departure from traditional cloud-centric project tools.
The core of Threlmark’s design is that all project data resides in plain JSON files on the user’s disk, with the on-disk layout serving as the API. This decision eliminates the need for a central server, enabling users to control, back up, and migrate their data easily. The system organizes data into a directory structure, with each project, card, and external suggestion stored as individual files, facilitating inspection, versioning, and sharing without vendor lock-in. The architecture employs atomic file writes through temporary files and rename operations, ensuring data integrity even during crashes or interruptions. Updates to project items are handled via read-merge-write cycles, which preserve existing data and support forward compatibility by maintaining unknown fields. This design allows external tools and AI agents to participate seamlessly by reading and writing files directly, without requiring special permissions or APIs.Disk is the contract: inside a local-first roadmap hub
A Next.js app on top of plain JSON files — no database, no cloud, no accounts. The key decision: the on-disk layout IS the API. Everything else cascades from taking that seriously.
There is no server-of-record — the files are the record
The UI and any external tool reach the same files through the same discipline. The data root defaults to ~/.threlmark — home-based, because it’s a shared hub every one of your apps points at.
Inspectable
Every artifact is a file you can cat, diff, grep, commit.
Portable · no lock-in
Back up with cp, sync with Dropbox / git, migrate trivially.
Interoperable
Any tool in any language joins by reading / writing files.
Restartable
No in-memory state to lose — stateless over the files.
Samsung T7 Portable SSD, 1TB External Solid State Drive, Speeds Up to 1,050MB/s, USB 3.2 Gen 2, Reliable Storage for Gaming, Students, Professionals, MU-PC1T0T/AM, Gray
- Designed for Creators: Fast speeds and durable for all tasks
- Fast Data Transfer: Up to 1,050MB/s read/write speeds
- High Capacity Storage: Up to 4TB for large files and games
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Two disciplined patterns instead of a database
“Just use files” is easy to get wrong. These two patterns — ported from a battle-tested sibling app — are what make file-based state sound rather than reckless.
Atomic writes
Write to a temp file in the same dir, then rename() over the target. Rename is atomic on one filesystem — a crash mid-write leaves the complete old file or the complete new one, never a half.
The board heals itself
A single roadmap.json array races when two tools write at once. One file per card makes writes collision-free. Lane order lives in board.json and reconciles on read.
board.json. It writes an item file — the board fixes itself on Threlmark’s next read. Unknown keys are preserved, so the contract is forward-compatible.The numbers can’t drift from the files
Anything computable from item state is computed — so the displayed numbers can never disagree with the underlying JSON. Priority is the clearest example: it’s calculated on read, never persisted.
priority — computed on read
Impact weighted heaviest; effort the only axis that subtracts. Reused verbatim from the original tool, so imported cards rank identically.
A handoff is a first-class flow event
The genuinely 2026-shaped part: most building is done by AI agents, so Threlmark closes the loop. Watch a card go from ranked to Done without anyone dragging it.
Handoff → report → self-move
The brief carries a reporting protocol. The agent reports through REST or the filesystem — and a done report moves the card itself.
POST /api/projects/:id/
items/:itemId/reportDirect call. Applied immediately.
drop reports/.json
→ ingested on read Robust even if the server’s down at finish time.
A small formula, and an honest hosting caveat
Because items are globally addressable (
Portfolio ranking — status-weighted
In-flight work floats to the top; bottlenecks cost the most, so blockers get nudged up.
Static read-only demo
Seeded data, writes to localStorage. Try-before-you-clone.
Personal Node instance
Password-gated, persistent backed-up THRELMARK_DATA_DIR.
Multi-tenant SaaS
Add accounts + per-tenant isolation. A separate build.
src/lib/*/store.ts is the natural seam — the same boundary that keeps the local tool simple is the one you’d extend for multi-tenancy. The architecture doesn’t fight that future; it just doesn’t pay for it until you need it.
Why a Disk-Based Approach Changes Project Management
Threlmark’s local-first, file-based architecture fundamentally shifts how project data is stored, accessed, and manipulated. By removing reliance on centralized servers and cloud services, it enhances data control, reduces lock-in, and simplifies backups and migrations. The approach also enables more reliable AI automation, as agents can operate directly on the same files used by humans, fostering more integrated workflows. For users and developers, this means more transparency, flexibility, and resilience in managing complex projects.
Background of Threlmark’s Design Philosophy
Traditional project management tools often depend on cloud servers and databases, which can introduce issues related to data lock-in, privacy, and dependency on external services. Threlmark’s approach draws inspiration from local-first principles and previous software that prioritizes file-based storage for safety and portability. Its architecture aligns with ongoing trends toward decentralization and user sovereignty over data, while also aiming to facilitate AI integration in a way that is transparent and controllable.
“The key decision was to make the disk the contract, so all tools and AI agents work directly with the same files, no databases needed.”
— Thorsten Meyer, creator of Threlmark
Remaining Questions About Threlmark’s Architecture
While the design principles are clear, it is not yet confirmed how well the system performs under heavy concurrency or with very large projects. Additionally, the extent of external tool and AI agent support beyond initial implementations remains to be seen, as does user adoption in diverse workflows. Details about integration with existing project ecosystems are still emerging.
Next Steps for Threlmark and Its Ecosystem
Threlmark plans to release more detailed documentation and developer tools to facilitate third-party integrations. Future updates may include enhancements to AI automation capabilities, performance optimizations, and broader community testing. Observers will watch for how well the architecture scales and how external tools adopt the file-based model.
Key Questions
How does Threlmark handle concurrency without a database?
It uses atomic file writes via temporary files and renaming, combined with read-merge-write cycles that preserve data integrity even when multiple tools access files simultaneously.
Can external tools participate in managing projects?
Yes, any tool that can read and write JSON files can participate, making the system highly interoperable and flexible.
What are the advantages of a disk-based, serverless architecture?
It offers greater data control, easier backups and migrations, and eliminates vendor lock-in, while supporting AI automation directly on project files.
Is this approach scalable for large teams or complex projects?
It remains to be tested at scale; initial indications suggest good performance for small to medium projects, but scalability details are still under evaluation.Source: ThorstenMeyerAI.com
