AIThis post was created with the assistance of artificial intelligence (AI).

TL;DR

Prime Big Deal Days · Oct 6–7Offer from Amazon

Get garden gear delivered free — and shop member deals

  • Fast, free delivery on millions of items
  • Access to Prime Big Deal Days deals on October 6–7
  • Prime Video, Amazon Music and more included
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

Threlmark has been presented as a roadmap tool built around a local JSON file rather than a hosted database or API. Its core claim is that teams, tools and agents can share the same roadmap by reading and writing one file, while users must still review automated changes.

Threlmark, a roadmap and scored kanban tool from Thorsten Meyer AI, is developed privately and is not publicly available. It has been presented as a tool built around a local JSON file that serves as the shared contract for the app, other tools and automated agents. The development matters because the product is being positioned against roadmap tools that keep plans inside hosted databases and APIs controlled by vendors.

The Thorsten Meyer AI dispatch says Threlmark stores the roadmap as a plain JSON file on the user’s disk. The kanban board is described as a view over that file, not the primary home of the data. According to the source material, any tool that can read or write JSON can act as a client for the roadmap.

The project is developed privately and is not publicly available. The dispatch frames the design around four words: “disk is the contract.” In practical terms, that means the contract is the file format and structure rather than a SaaS database, SDK, webhook system or private API.

Threlmark is also described as a scored kanban system. Each roadmap item carries a score, and the board ranks work by those scores. The source says this connects Threlmark to IdeaClyst, another product in the same operator portfolio: IdeaClyst produces verdicts, while Threlmark turns those verdicts into ordered roadmap items.

Built in Public · Day 7 / 19 ThorstenMeyerAI.com · the operator portfolio
The Decision Layer · Day 07 Dispatch

Threlmark — disk is the contract

The roadmap is a plain JSON file on your disk. The board is just a view over it — and your tools and your agents read and write the same file directly.

01 One file. Everything reads & writes it.
Threlmark UI+ your tools
⇄
{ }
roadmap.json
the contract
⇄
Agentsread · act · write
read → act → write · no API, no lock-in — a plain file any program can honor
Backlog
Bulk CSV importer
score 49
Niche export format
score 58
Scored← council
Build: validated idea X
score 87
Build: validated idea Y
score 74
Doing
Ship feature Z
score 91
Done
Launch W
✓ shipped
1 filethe whole roadmap, on disk scoredevery item ranked JSONfile-based · agent-readable
02 Why a file beats a database here
JSON
the contract is a file format, not a vendor — anything that reads JSON is a client.
own it
a text file you own, that outlives any tool — no API, no rate limit, no lock-in.
agent-native
the roadmap is a shared workspace — humans and agents write to the same file.
03 The thesis the whole series inherits
01
Local-first
The roadmap is literally a file on your machine — not a row in someone else’s database.
02
Provider-agnostic
The contract is a format, not a vendor. Any agent, any tool that speaks JSON is first-class.
03
Non-developer build
Radical simplicity by design — the least lock-in-prone thing there is: a plain file.
04
Edit by subtraction
Scoring forces ranking. A board where everything is “high priority” has no priorities.
04 The operator constellation
18 products · one foundation
Today: Threlmark lit — where the council’s verdicts become an ordered plan. IdeaClyst → Threlmark.
Content
DojoClaw
RoundupForge
Stenvrik
ChannelHelm
IdeaNavigator
Decision
IdeaClyst
Threlmark
Outcome-First
Platform
Grimfaste
Delvasta
Open / Reg
Glasspane
QAtrial
Markets
Polybot
TradingAgents
Defense / Intel
Argus
VigilSAR
VigilSAR-Bench
Diagnostic
World Model Readiness
Local-first · Provider-agnostic foundation

Independent commentary, produced with AI assistance under human editorial oversight. The views are the author’s own and may change. Threlmark is developed privately and is not publicly available. Automated agents that read and write the roadmap file may introduce errors — treat agent writes as changes to review, not facts to trust. Product and company names are trademarks of their respective owners; mention does not imply endorsement.

ThorstenMeyerAI.com · Built in Public · Day 7 of 19 · © 2026 Thorsten Meyer

Local Files Challenge SaaS Lock-In

The release points to a broader dispute in software tooling: where operational data should live. Many roadmap products store planning data in vendor-controlled systems. Threlmark’s approach makes the file itself the integration point, which could make it easier for users to inspect, back up, version, move or edit their roadmap outside the original app.

For readers building with AI agents, the agent angle is the clearest use case. The source material says humans and agents can read and write the same roadmap file directly. If that works as described, an automated agent could add, rank or update work without a custom integration layer. That could reduce setup friction for local-first workflows, though it also creates review duties for users.

The scoring model also gives the tool a sharper product stance than a basic task board. By requiring each item to carry a score, Threlmark is meant to make trade-offs visible. The dispatch says a roadmap where everything is high priority has no priorities, and the score is the mechanism meant to force ranking.

Amazon

JSON file editor for project management

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Day Seven In The Portfolio

The announcement appears in the Built in Public series as Day 7 of 19. The source frames Threlmark as part of an 18-product “operator constellation,” with the current connection described as IdeaClyst → Threlmark. In that chain, IdeaClyst is presented as the decision source and Threlmark as the place where those decisions become scheduled work.

The dispatch also places Threlmark inside three recurring themes for the series: local-first data, provider-agnostic tooling and low lock-in. The relevant background is narrow: this is not being presented as only a kanban interface, but as a data ownership decision. The roadmap is meant to survive the tool because it is stored as a text file in a known shape.

The source includes cautionary language. It says the project is provided “as is” without warranty, and that automated agents writing to the roadmap may introduce errors. That means agent edits are being framed as changes to inspect, not facts users should accept without review.

“disk is the contract”

— Thorsten Meyer AI dispatch

Amazon

Kanban board software with local file support

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Release Details Still Limited

The source material confirms the product thesis, license claim, file-based model, scored-kanban concept and relationship to IdeaClyst. It does not provide independent usage metrics, installation steps, repository activity, supported schema details, compatibility guarantees or a changelog for the current build.

It is also not clear from the provided material how Threlmark handles simultaneous edits, schema migration, file corruption, conflict resolution, authentication around shared files or multi-user workflows. Those details matter for teams that might want to use a local JSON roadmap beyond a single-machine setup.

The dispatch claims that the local JSON approach avoids API limits and vendor lock-in. That is consistent with the described architecture, but the real-world durability of the format depends on whether the schema remains stable and documented over time.

Amazon

roadmap planning tools without cloud API

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Schema And Adoption Tests

The next milestone for readers is practical verification: reviewing the project, checking the license, inspecting the roadmap JSON shape and testing whether external tools and agents can safely read and write the file. Teams considering the tool should also decide how they would version the file and review automated edits before treating them as accepted roadmap changes.

The Built in Public series is scheduled to continue beyond Day 7, so further details may come through later dispatches. For now, the confirmed development is a product announcement centered on local, file-based roadmap data rather than a hosted planning database.

Amazon

local JSON-based project tracking tool

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Key Questions

What is Threlmark?

Threlmark is described as a roadmap and scored kanban tool from Thorsten Meyer AI that is developed privately and is not publicly available.

What does “disk is the contract” mean?

It means the roadmap data is stored as a plain JSON file on the user’s disk, and that file structure is the integration contract for the app, other tools and agents.

How is Threlmark different from typical roadmap tools?

The source contrasts Threlmark with tools that keep roadmap data in hosted databases behind vendor APIs. Threlmark’s roadmap is presented as a local file that users can own, inspect and edit with any compatible JSON tool.

What role do AI agents play?

The dispatch says agents can read and write the same roadmap file as humans and the Threlmark UI. The source also warns that automated writes may introduce errors, so those changes should be reviewed.

What is still unknown about Threlmark?

The provided material does not specify repository activity, current installation steps, full schema rules, multi-user handling, conflict management or adoption data.

Source: Thorsten Meyer AI

FALL YARD WORK

Fall yard work Picks

As an affiliate, we earn on qualifying purchases.

You May Also Like

The Door: Why the Interface Is Worth More Than the Model

Thorsten Meyer AI says a reported SpaceX-Cursor deal shows user-facing AI interfaces can steer models, data and demand.

Why AI’s Management Flaws Surface Even When It Gets The Right Answer

Firmulate says five frontier AI models spotted every crisis, but only two completed a €55,000 software deal.

Top 9 AI-Powered Mobile Workstation Laptops For Professionals In 2026

A 2026 comparison ranks Dell’s Precision 7680 first, while Lenovo and HP models target portable, visual and enterprise workloads.

The Model Is Only 10%: The Real Lesson of the New SDLC

A May 2026 Google paper says AI coding success depends less on models than on tests, tools, context and oversight.