TL;DR
Thorsten Meyer AI has presented Threlmark, an MIT-licensed roadmap and scored kanban tool built around a local JSON file. The project’s core claim is that the roadmap file, not a hosted database or API, is the shared contract for the UI, external tools and AI agents.
Thorsten Meyer AI has presented Threlmark, an open-source roadmap and scored kanban tool that stores the plan as a local JSON file rather than in a hosted product database, making the file itself the shared contract for people, tools and AI agents.
The source material describes Threlmark as MIT-licensed software available at threlmark.com. Its main product decision is that the roadmap is a plain JSON file on the user’s disk. The kanban board is described as a view over that file, while other tools and automated agents can read from and write to the same file directly.
Thorsten Meyer AI frames that design as a rejection of roadmap tools that keep plans inside vendor databases and expose them through hosted APIs. The dispatch says Threlmark has no SaaS API layer for the roadmap contract; instead, the file format is the integration point. That claim matters most for users who want their operational plan to remain readable outside one product.
The product is also described as a scored kanban board. Items carry priority scores, so the roadmap can rank work instead of treating all cards as equal. The source links Threlmark to IdeaClyst in the wider Thorsten Meyer AI product portfolio, saying IdeaClyst verdicts can become scored items in Threlmark.
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.
Independent commentary, produced with AI assistance under human editorial oversight. The views are the author’s own and may change. Threlmark is open source under MIT, provided “as is” without warranty; see the repository LICENSE. 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.
Local Files Reduce Roadmap Lock-In
If Threlmark works as described, its importance is less about another kanban interface and more about ownership of the plan. A roadmap often becomes the working memory for a product team or solo operator. When that data is locked inside a service, access can depend on pricing, API availability, export quality and the vendor’s future.
A local JSON file changes that trade-off. JSON can be read by common developer tools, edited by scripts and tracked in version control. The source argues that this makes the roadmap more durable and easier for AI agents to use because an agent can work against the same file as the human-facing board.
Those benefits remain claims from the project rather than evidence of broad use. The confirmed development is the product direction and architecture described in the dispatch. The practical impact will depend on the stability of Threlmark’s file format, how safely write operations are handled and whether users find file-based roadmaps easier than hosted project tools.
local JSON file project management tool
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Day Seven In Product Series
Threlmark is presented as Day 7 of a 19-part Built in Public series from Thorsten Meyer AI. The dispatch places it within what the source calls an operator portfolio of 18 products built on a local-first and provider-agnostic foundation.
The immediate link in that portfolio is IdeaClyst to Threlmark. In the source’s framing, IdeaClyst produces verdicts about which ideas are worth pursuing, while Threlmark turns those verdicts into an ordered plan. The roadmap examples in the source show backlog, doing and done columns, with scores attached to proposed work.
The source also includes limits and cautions. It says Threlmark is provided under the MIT license and provided as is, without warranty. It also warns that automated agents writing to the roadmap may introduce errors, so their changes should be reviewed.
“The roadmap is a plain JSON file on your disk.”
— Thorsten Meyer AI dispatch
Adoption And Safety Still Open
Several details are not established in the source material. It does not provide adoption numbers, active user counts, a repository URL, version number or a calendar release date. It also does not show how conflicts are handled when several tools or agents edit the roadmap file at the same time.
It is also not clear how stable the JSON schema will be, how migrations will work if the format changes, or what validation tools are included to catch malformed agent writes. Those questions matter because the same file-based design that makes Threlmark easy to inspect can also expose users to accidental edits if review workflows are weak.
Schema Stability Comes Next
The next test for Threlmark is whether the project can turn its file-first design into a stable working contract. Readers should watch for the public repository, schema documentation, examples of agent writes, validation rules and guidance for version control workflows.
Further Built in Public entries may also clarify how Threlmark connects with the rest of the Thorsten Meyer AI portfolio, including IdeaClyst. For now, the confirmed story is the product direction: Threlmark makes the roadmap a local JSON file and treats the board, tools and agents as clients of that file.
Key Questions
What is Threlmark?
Threlmark is described by Thorsten Meyer AI as an open-source roadmap and scored kanban tool. Its roadmap data lives in a local JSON file, while the board acts as a visual view over that file.
What does disk is the contract mean?
It means the shared agreement between Threlmark, other tools and AI agents is the file on disk. The source says any program that can read and write JSON can work with the roadmap format.
Is Threlmark a SaaS roadmap tool?
The source frames Threlmark differently from SaaS roadmap tools. It says the roadmap is stored as a local file rather than held behind a hosted database and API.
Can AI agents edit the roadmap?
According to the source material, agents can read from and write to the roadmap file. The same source warns that agent writes may contain errors and should be reviewed.
What is still unknown?
The source does not confirm user adoption, release version, schema governance, conflict handling or detailed validation features. Those details will shape how practical the approach is outside the demo and dispatch.
Source: Thorsten Meyer AI