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

Ask anyone who makes content for a living where a particular thing is — the brief, the latest draft, the asset, the client’s feedback, the prompt they used last time — and watch them pause. The brief is in a doc. The calendar is a spreadsheet. The assets are in a drive folder named something hopeful like final_v3. The feedback is buried in a chat thread. The prompt is in a note they’ll have to rewrite because they can’t find it. The client-facing version is in yet another tool entirely. None of these things know about each other. The work isn’t lost, exactly; it’s just scattered, and scatter has a cost that compounds quietly every week — the time spent hunting, copying, asking “which version is current,” and rebuilding things that already existed.

Buying for a business?Offer from Amazon

Get business pricing on tech for your team

  • Business-only prices and quantity discounts
  • Tax-exempt purchasing
  • Multiple users, one account, clear invoices
As an affiliate, we earn on qualifying purchases.

Thrymvault is built against that scatter. It’s a private, self-hosted content workspace that pulls the whole workflow — ideas, drafts, assets, clients, feedback, repeatable AI prompts — into one place where the pieces actually know about each other. Rich pages and flexible databases, public portals and threaded comments, a file library and full-text search, all in a workspace you run yourself. The promise isn’t a new app to check. It’s the opposite: one place, so there are fewer places.

Thrymvault · A System Around Your Content · Built in Public Spotlight
Built in Public · Spotlight · Thrymvault ThorstenMeyerAI.com · the operator portfolio
Self-hosted content workspace · pages + databases + portals

A System Around Your Content

One self-hosted workspace where ideas, drafts, assets, clients, feedback, and reusable AI prompts finally know about each other — instead of scattered across notes, sheets, folders, and chat threads.

01 Documents and databases, one room
one content database · four saved views · zero duplicated rows
Queue
Board
Calendar
Archive

Typed properties, relations, and saved views mean the same records become a writing queue, a kanban board, a calendar, or a searchable archive — and each record carries a rich-text body, so the plan and the draft live together.

02 The daily loop — connected, not scattered
01
Capture
An idea lands in the content database before it gets lost.
02
Enrich
Research, files, and draft notes go in the record body.
03
Progress
Move it through a board as it advances.
04
AI run
Saved prompts generate outlines, summaries, variants.
05
Review
Comments and @mentions, attached to the work.
06
Schedule
Drop it onto a calendar view.
07
Share
Project it through a client or stakeholder portal.
08
Search
Find it again when the next project rhymes.
03 Portals — the polished pieces, not the messy middle
★ read-only projection · property-level whitelist
Clients see the finished surface. Your internal notes, hidden fields, comments, and private records never leave the workspace.
Private workspace
Published calendar
Deliverable status
Internal notes
Hidden properties
Comments & records
whitelist
+ token
+ passphrase
→
Public portal
Published calendar
Deliverable status
— nothing else —
04 The part that makes it yours
Self-hosted
Built on a self-hosted Convex backend — you run the workspace, you keep the data.
Real access
Roles, item-level shares, server-side authorization, and scoped guest access.
LAN-first
Local-network deployment as a first-class option, not an afterthought.
Exit kept open
Start self-hosted, move to hosted later via env changes — not a rebuild.
05 Honestly labeled — what this is
the thesis of the tool, not a claim that every surface is finished
  • This is the capability set. Drawn from Thrymvault’s own product documentation — what the workspace is for and how its pieces fit.
  • Early-stage, in active build. Some surfaces are more settled than others; treat described capabilities as design, not a finished-product guarantee.
  • No deploy-and-verify story yet. Unlike the shipped products in this series, there’s no public-launch writeup attached here — when there is, it gets the same treatment.
  • The promise is “lose less.” Not “do more” — less time hunting, copying, asking, and rebuilding, because the pieces share one roof you own.

Independent commentary, produced with AI assistance under human editorial oversight. The views are the author’s own and may change. This is not business, financial, legal, or technical advice. Thrymvault is an early-stage, self-hosted product in active development; described capabilities reflect its design and may change. Product, model, and company names are trademarks of their respective owners; mention does not imply endorsement.

ThorstenMeyerAI.com · Built in Public · Spotlight · Thrymvault · © 2026 Thorsten Meyer

Documents and databases, finally in the same room

The core move is deceptively simple. Most tools force a choice: a documents app for the freeform stuff (briefs, scripts, research, long-form drafts) or a database app for the structured stuff (calendars, trackers, CRMs, pipelines). You end up with both, and the seam between them is where work falls through.

Thrymvault refuses the choice. A page can hold a long-form draft and the plan for it. A database has typed properties, relations, and saved views — so the same content records can be looked at as a writing queue, a kanban production board, a calendar, a sponsor tracker, or a searchable archive, without duplicating a single row. A creator’s content database might carry ideas, videos, posts, newsletters, episodes, hooks, thumbnails, scripts, publishing dates, and performance notes — each record also having a rich-text body, so the plan and the actual draft live together instead of in two tools that have to be kept in sync by hand.

That combination — the freedom of documents with the structure of databases — is what turns a pile of notes into something that behaves like a content operating system: every piece of work gets a clear place, a status, an owner, the surrounding context, and a next step. The structure isn’t bureaucracy. It’s the thing that lets an idea move from backlog to research to draft to scheduled to published to repurposed without anyone losing track of where it is.

Amazon

self-hosted content management system

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

An AI layer built around saved prompts, not a chat box

Most AI features bolted onto productivity tools are a blank chat box: useful once, then you’re retyping the same instructions next week and getting slightly different output each time. Thrymvault’s AI layer is built around saved prompts instead — a small library of proven workflows you run against pages, records, selected text, or selected rows.

Turn rough notes into a video outline. Summarize a block of research. Generate title options. Rewrite a paragraph for a different channel. Extract a newsletter from a long article. Spin a core idea into social variants. The point isn’t that any one of these is novel; it’s consistency. You build the prompt once, get it right, and reuse it, rather than reinventing it under deadline. And because runs can go in bulk across many selected rows, the repetitive work — the same transformation applied to forty records — stops being forty manual tasks. The model does the typing; the saved prompt encodes the judgment about how.

Amazon

digital asset management software

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Portals: publish the polished pieces, not the messy middle

If there’s one feature that earns Thrymvault its keep for client work, it’s portals, and it’s worth being precise about what they do because the idea is genuinely useful.

A portal is a read-only projection of selected pages and database views, shared through a tokenized public link — with optional passphrase protection, branding, and property-level whitelisting. The whitelisting is the part that matters. You choose exactly which fields a recipient sees. So a client gets a clean content calendar, a campaign status dashboard, or a deliverable library, while the internal notes, hidden properties, private records, comments, and workspace navigation stay entirely out of view. They see the polished pieces; they never see the messy middle where the actual work happens.

That solves a problem every agency and consultant knows intimately: the tension between giving a client visibility and giving them access. Portals collapse it. The same workspace that runs your internal chaos can publish a stakeholder-ready surface without you maintaining a second, sanitized copy of everything — and without a client login into your private workspace. Client content calendars, campaign readouts, public resource hubs, product catalogs, editorial review portals: all projections of work that already exists, never duplicates of it.

Amazon

private knowledge base software

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Feedback that stays attached to the work

The collaboration model follows the same logic as everything else — keep the thing connected to its context. Threaded comments, replies, resolution, and @mentions live on the exact page or record they’re about, so “who has the latest feedback?” and “where was that note?” stop being questions. A notification bell pulls people back to the comment that needs them, which matters most in the review-heavy loops — editing, approvals, client feedback, publishing coordination — where feedback scattered across channels is precisely what slows everything down.

Underneath that, the workspace has real access structure: membership and roles, item-level sharing, server-side authorization, and guests who can be given specific access without the whole workspace opening up. Public portals remain a separate surface from the authenticated workspace, which is what makes “share this with a client” safe by construction rather than by remembering to be careful.

Amazon

project management database tools

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

An archive that compounds

Every creator and team eventually accumulates a private body of knowledge: old hooks that worked, proven formats, research, sponsorship notes, audience questions, lessons learned. In a scattered setup that knowledge is effectively write-only — you make it and never find it again. Thrymvault makes it searchable across pages, databases, and records from a single command palette, so yesterday’s work becomes retrievable leverage for tomorrow’s. The value isn’t the search feature in isolation; it’s that the archive exists in one place to be searched at all. Consolidation is what makes retrieval possible.

The part that makes it yours

Here is where Thrymvault sits squarely inside the rest of this portfolio. It’s built on a self-hosted Convex backend, designed around workspace membership, roles, item-level sharing, and server-side authorization, with LAN deployment as a first-class option rather than an afterthought. For anyone who cares about data ownership — agencies handling client material, businesses with confidentiality requirements, teams that want a private alternative to a patchwork of cloud tools — that’s the difference between adopting a workspace and renting one.

And it doesn’t trap you in the choice: a workspace can start self-hosted on hardware you already own and move to hosted infrastructure later through environment-level changes rather than a total rebuild. Own it now, keep the option to grow. That’s the same stance the whole portfolio keeps taking — own your ground, refuse the lock-in, keep the exits open — applied here to the place your content actually lives.

Who it’s for, and when it fits

Thrymvault is a strong fit for independent creators running a serious publishing workflow, agencies producing content for multiple clients, small businesses that need content, light CRM, and knowledge management in one place, founders building repeatable marketing operations, consultants sharing polished resources with clients, and internal teams that want a private, self-hosted alternative to a stack of disconnected tools.

It’s most valuable precisely when the work has both a creative and an operational side — ideas and drafts, but also assets, approvals, dates, clients, and deliverables. A pure notes app handles the creative side and drops the operational one; a pure database handles the operational side and strangles the creative one. The whole reason to combine them is that real content work is both at once.

An honest note on what this is

In keeping with how this series is written, the plain version: this spotlight describes Thrymvault’s design and capability set — what the workspace is for and how its pieces fit together — drawn from the product’s own capabilities documentation. Thrymvault is an early-stage, self-hosted build, and some capabilities described here are more settled than others. There’s no deploy-and-verify story attached to this piece the way there is for the products in this series that already shipped a public face; when there is, it’ll get the same treatment. Read this as the thesis of the tool, honestly labeled as such, not a claim that every surface is finished.

Why it belongs

Thrymvault doesn’t just store content; it builds a system around content. That’s the actual promise, and it’s a quieter one than most software makes. Not “do more,” but “lose less” — less time hunting, copying, asking, and rebuilding, because the brief, the draft, the asset, the feedback, the prompt, and the client-facing view finally live in one workspace that you own and that knows how its own pieces connect.

For a creator, that’s a place where ideas reliably become publishable assets. For a business, it’s a workspace where marketing, operations, client communication, and AI-assisted work can share a roof. For everyone, it’s the calm that comes from one place instead of seven. In a portfolio built on owning your ground and removing more than you add, a tool whose entire job is to consolidate the scatter and keep it yours is right at home.


Independent commentary, produced with AI assistance under human editorial oversight; the views are the author’s own and may change. This is not business, financial, legal, or technical advice. Thrymvault is an early-stage, self-hosted product in active development; described capabilities reflect its design and may change, and not every surface should be assumed finished. Product, model, and company names are trademarks of their respective owners; mention does not imply affiliation, sponsorship, or endorsement. © 2026 Thorsten Meyer · Powered by Thorsten Meyer AI. See Imprint/Impressum and Privacy Policy.

FALL

Fall Picks

As an affiliate, we earn on qualifying purchases.

You May Also Like

The Google I/O 2026 Preview: What May 19-20 Will Reveal About Google’s Agentic Bet

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

NicheCommand: A Firehose Becomes a Shortlist

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

The Great Rewiring: A Thorsten Meyer AI Field Guide to the Post‑Labor Economy

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

OpenAI Cuts Off Cursor: The Developers Are the Collateral

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