🔍 Read the full analysis: Making A Shared Desk Work For AI Agents on ThorstenMeyerAI.com
Get hardware and tech essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Thorsten Meyer says his StrongMocha News Group newsroom is building Thrymvault as a shared workspace for human editors and AI agents, replacing chat-only coordination with story boards, attached reports and traceable changes. The system is self-hosted and restricts agents to approved applications, but some planned safeguards, including a standby machine and hourly backups, are not yet live.
Thorsten Meyer has described a shared workspace called Thrymvault for coordinating AI agents and editors across his StrongMocha News Group sites, after a newsroom workflow using chat and a code-oriented task tracker left story status and editorial changes difficult to follow. The system is designed to put articles, agent reports and editor instructions in one place, with changes attributed to approved applications; Meyer says some planned infrastructure is not yet live.
Meyer wrote that his team used a chat window and a task tracker built for code work to guide nine articles through the newsroom in one day this month. Agents gathered sources, drafted copy, built graphics and prepared drafts. But he said he had to ask for updates, send headline changes through chat and could not share a completed explainer page as a link. He described those limits as evidence that an AI-run newsroom needs a shared desk rather than conversations alone.
Thrymvault is a workspace for pages, databases and boards that Meyer’s group built and hosts on its own machines. According to his account, agents can connect through an MCP server with 20 tools, a REST API or a command-line tool. Its plain-text guide and machine-readable manifest are generated from the same tool catalogue as the server, which Meyer says keeps agent instructions aligned with available tools.
The proposed editorial workflow tracks a story from a topic card to source research, drafting, editor feedback and publication. Agents are expected to update their status and add reports beside the relevant card. An editor can move a card or leave a comment that mentions an agent’s application. Meyer says agent changes carry the name of the approved application, while agents cannot delete content. Moving a card is an agreement the agents follow, he notes, rather than a software switch that automatically starts work.
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 Workflow for AI Newsrooms
The project addresses a practical accountability problem: when multiple agents work on sources, copy and graphics, a chat thread can make it hard to see which version is current, who requested a revision or what instruction governed the work. Attaching status, reports and comments to a story card could give editors a more persistent record and a clearer place to intervene.
Meyer also presents the design as a control measure, not only a productivity tool. Agent access requires an owner-approved application, and keys operate with the rights of the person who created them, within one workspace and no more rights than the application allows, according to his description. These limits may help organizations review agent activity, though the article does not provide an independent security assessment or evidence of results at scale.
For readers, the relevance is indirect but concrete: a documented chain of editorial changes could make it easier for a newsroom to identify how a draft was produced and correct mistakes. That benefit depends on whether the records are complete and whether editors consistently use them. The workspace itself does not establish that an article is accurate.
Products Worth Considering
Trezor Safe 7 Crypto Hardware Wallet with Bluetooth and Color Touchscreen
Ledger Flex Cryptocurrency Wallet with 2.8-Inch Touchscreen
Apple MacBook Pro 16.3-inch with PROLED XDR Display, Silver
As an affiliate, we earn on qualifying purchases.
From Chat Threads to Story Cards
Meyer’s account begins with a workflow in which agents and an editor coordinated through chat and a task tracker intended for software work. He says that setup could handle research and production tasks, but story status had to be requested and instructions could scroll out of view. In his view, those weaknesses matter because agents may work in parallel and need written instructions to retain decisions across tasks.
Thrymvault is described as having a familiar pages-and-databases format, but the specific newsroom design adds a shared editorial board. A topic moves through stages such as “Topic,” “Kit” and “Draft.” The editor chooses which topics proceed; agents then prepare source material and drafts, while unresolved questions and reports stay linked to the story.
Decisions that affect future work, such as a headline rule, are intended to be recorded in a separate database with the wording, date, scope and link to the rule it changes. Meyer says the rule remains where agents read it, while the database records the decision behind it. The account is a description of the group’s own system and plans, not a comparative test against other newsroom platforms.
“That is the point where an AI-run newsroom outgrows chat.”
— Thorsten Meyer, describing the newsroom workflow
Rollout and Safeguards Still Developing
Meyer distinguishes the system’s design from what is already operating, but the available account does not give a detailed feature-by-feature status list. He says a second machine on standby and hourly backups are not live yet. It is also unclear how much of the described story workflow is already in routine newsroom use, how many agents or applications currently use it, and whether the planned controls have been tested under heavy use.
The article does not report independent audits, incident data, performance measurements or a comparison with other tools. It also does not specify how long activity records are retained, how access is revoked in every case, or how editors verify agent-generated sources and claims. The described permissions and attribution are Meyer’s account of Thrymvault’s design and documentation, not an outside certification of security or editorial quality.
Backups and Newsroom Adoption
Meyer identifies the next infrastructure step as adding a standby machine and hourly backups. He does not give a date for that work. The broader rollout is also not accompanied by a published timetable or a list of milestones in the source material.
For the newsroom workflow, the next test is whether editors and agents consistently use the cards, comments and decision records as stories move toward publication. Meyer says the workspace is intended to leave a trace of card movements and agent changes, but he does not yet report measured effects on accuracy, turnaround time or correction rates. Those outcomes will depend on how the system is implemented and reviewed in daily use.
Key Questions
What is Thrymvault?
Thrymvault is a workspace for pages, databases and boards that Thorsten Meyer says his group built and hosts on its own machines. It is designed to let newsroom editors and AI agents coordinate work around shared story records.
How would an AI agent use the newsroom desk?
According to Meyer, agents can work through an MCP server, REST API or command-line tool. In the proposed workflow, an agent researches a story, drafts it, updates its status and responds to editor comments attached to the story card.
What controls does Meyer say are in place?
Meyer says agents can enter only through owner-approved applications, changes are attributed to the application used, and agents cannot delete content. He also says a key is limited to one workspace and cannot grant more rights than its application has.
Is the backup system already operating?
No. Meyer says the planned standby machine and hourly backups are a next step and are not live yet.
Does the workspace verify that AI-written stories are accurate?
The source describes workflow tracking and access controls, not an accuracy guarantee. Editors would still need to check sources, claims and drafts; Meyer reports no independent evaluation of Thrymvault’s effect on accuracy.
Source: ThorstenMeyerAI.com
