🔍 Read the full analysis: What I Learned From Giving My AI Agents A Shared Desk on ThorstenMeyerAI.com
Prime made for students and young adults
- Fast, free delivery for dorm and study essentials
- Prime Video and Amazon Music included
- Member-only deals
TL;DR
Thorsten Meyer says his AI-run newsroom is moving work from chat and a code-focused task tracker into Thrymvault, a shared workspace built for pages, databases and boards. The system is designed to track story status and changes while limiting agents to approved applications; some planned infrastructure is not yet live.
Thorsten Meyer says his AI-run newsroom is shifting its editorial coordination from chat and a code-focused task tracker to Thrymvault, a shared workspace intended to show story status, collect instructions and record changes made by agents. The project matters because it sets out a more traceable way to manage AI-assisted reporting, but Meyer’s account describes a mix of current features and plans still being rolled out.
Meyer wrote that, on one day earlier in November, his team steered nine articles through a chat window and a task tracker built for code work. AI agents gathered sources, drafted material and prepared graphics, he said. But he had to ask for updates, issue headline changes in chat and rely on the intended agent seeing them. He also said he could not share a finished explainer page as a link.
Thrymvault is described as a workspace for pages, databases and boards, built and hosted on Meyer’s own machines. Agents can connect through an MCP server with 20 tools, a REST API or a command-line tool. According to the project’s guide and integration documentation, access is through applications approved by a workspace owner; changes are attributed to the application, and agents cannot delete content.
In Meyer’s proposed newsroom workflow, a story moves across a board from topic selection to source gathering, drafting and publication. Agents update a story card with status, draft links and open questions; editors can leave comments and mention an agent application to request changes. The board is intended to retain the published link and a record of decisions. Meyer said moving a card signals an agreement for agents to follow; it is not itself a software switch that starts a task.
The shared desk · animated
How a story moves through the shared desk
One card per story. The editor moves it and asks questions in its thread; the agent does the work, keeps its status current and answers there. Every agent change is signed with its application.
Editorial board
Thread on the card
Editor@Editorial agent shorter title, no colon
AgentDone: title shortened, draft updated.
via Editorial agent
Signed history
- Moved to Kit · by the editor
- Status and draft link updated · via Editorial agent
- Comment mentions Editorial agent · by the editor
- Reply in the same thread · via Editorial agent
- Approved and scheduled · by the editor
- Published link added · via Editorial agent
Agents add and edit. They cannot delete anything.
Source: Thrymvault agent integration guide and rollout plan, November 2026. Shows the desk as designed for StrongMocha News Group; the board templates and the pick-up of mentions are still being rolled out.
Animation: Thorsten Meyer AI
A Traceable Desk for AI Work
The project addresses an operational problem for newsrooms using multiple AI agents: work can happen in parallel, while instructions and status updates can become difficult to track in a stream of chat messages. A shared board can give editors a single place to see what is underway and attach feedback to the relevant story.
Recording which approved application made a change could also help a newsroom review and correct work. Meyer argues that a record of who drafted, requested or changed material makes errors easier to investigate. That is a rationale for the system, not evidence that it has already improved accuracy or reduced errors. The account provides no independent evaluation of those outcomes.
For readers, the underlying issue is accountability: AI-generated or AI-edited material needs a workable review trail if editors are to understand how it reached publication. A shared workspace may help organize that process, but it does not by itself establish that sources were checked correctly or that published reporting is accurate.
Popular right now
Ninja Foodi 10-Qt DualZone XL Air Fryer, 6-in-1, Gray (DZ401)
Dyson V8 Motorhead Origin Cordless Vacuum
Dyson Supersonic Travel Hair Dryer
As an affiliate, we earn on qualifying purchases.
From Chat to a Newsroom Board
Meyer’s account is based on the needs of StrongMocha News Group, his network of niche and news sites, which he says operates with AI agents. He contrasts one-off chat conversations with a newsroom’s ongoing work on the same story, sources, graphics and deadlines. In his description, chat messages can scroll out of view, and parallel agents may interpret instructions differently.
Thrymvault is presented as a tool Meyer’s team built for its own use, with a general workspace format he compares to Notion. Its agent instructions are published as a plain-text guide and a machine-readable manifest, generated from the same tool catalogue as the server. Meyer says this is intended to keep agent instructions aligned with the tools available. The description and feature claims come from the project’s own materials, rather than an independent technical review.
“That is the point where an AI-run newsroom outgrows chat.”
— Thorsten Meyer, writing on ThorstenMeyerAI.com
Features Still Being Rolled Out
Meyer distinguishes the system’s design from what is already running, but the account does not give a full feature-by-feature deployment status or independent verification. The standby machine and hourly backups are explicitly described as a next step and not yet live. The article also does not report results from a sustained newsroom trial, such as how often agents miss comments or whether the audit trail has helped resolve a published error.
Other practical questions remain open: how access keys are revoked, how the workspace handles outages, and how editors verify agent-generated sourcing and drafts. Meyer says a key acts with the rights of the person who created it, within one workspace and with no more rights than its application allows, but the supplied account does not detail every permission or security control. No outside assessment of the platform’s security or reliability is provided.
Deployment and Backup Plans
Meyer says the newsroom is still rolling out Thrymvault’s workflow and identifies a second machine on standby, alongside hourly backups, as a future step. His account does not set a date for that deployment or announce a broader public launch. The next evidence to watch for is whether those backup arrangements go live and how the team reports on the system’s use in actual publishing work.
For now, the described workflow ends when a story is scheduled in the publishing system and goes live, with the published link added to its board card. Meyer’s account lays out how the desk is meant to operate; whether it delivers stronger editorial oversight in practice remains to be demonstrated.
Key Questions
What is Thrymvault?
Thrymvault is a workspace for pages, databases and boards that Meyer says his team built and hosts on its own machines. It is designed to let newsroom staff and approved AI-agent applications work around shared story records.
How do AI agents access the workspace?
Meyer’s documentation lists an MCP server with 20 tools, a REST API and a command-line tool. Access is limited to applications approved by the workspace owner, according to the project’s description.
Can agents delete newsroom content?
Meyer says agents cannot delete anything. Changes made through an agent are attributed to the approved application that made them, though the account does not provide an independent audit of those controls.
Is the backup system already running?
No. Meyer describes a second machine on standby and hourly backups as the next step, saying they are not live yet.
Has Thrymvault been shown to improve reporting accuracy?
The account does not provide data showing that Thrymvault improves accuracy, prevents errors or speeds publication. Meyer describes its intended tracking and review features, but their effects in newsroom use remain unclear.
Source: ThorstenMeyerAI.com
