Questions, answered.

A workspace for AI assets: prompts, agents, workflows, MCP tools and skills. You compose them from typed pieces, version them, test them with evals and share them. Your own AI runs them.

Yours: bring your own AI. Connect Claude, ChatGPT, Grok or any MCP assistant and it reads, runs and improves your prompts on its own subscription. Composing, versioning, sharing and publishing are done in the app.

It is in pre-beta: invite-only, with payments in test mode. The parts described on this site are built and in the app. We let people in a few at a time, so request access and we write when your place is ready.

Five kinds: prompts, agents, workflows, MCP tools and skills. A prompt can also be saved as a response (an example of the answer you want) or a context. They share tags, collections, history and sharing.

Yes. Search covers everything you own, by title, tag or text, across all kinds of asset.

A named shelf. It can hold any kind of asset, and an asset can be on several shelves. Taking an item off a shelf never touches the asset itself.

Only your workspace, by default. Others see an asset only when you share it by link (anyone with the link) or publish it on the marketplace. You can set a shared prompt back to private and the link stops working.

Write it in double braces, like {{customer_name}}. When you render the prompt you give the values; the render lists any variable you did not fill.

The render still happens, and it tells you which variables are missing, so you or your AI can ask for them. It also tells you what breaks the prompt’s type.

Eleven: Markdown, text, XML, JSON, YAML, HTML, CSV, TOML, JSONL, chat messages and Prompty.

No. Rendering is plain code: library pieces and nested prompts are resolved, wrappers applied and variables filled. The same input always gives the same text.

Prompts, workflows and agents. Each has commits, branches, tags and releases, and merge.

Merging a branch into the current one returns the conflicts so you can resolve them. Nothing is overwritten silently.

Create a branch from the older commit and check it out. The newer history stays where it is.

No. An eval scores every case from 0 to 1 and compares the result against a threshold you set, 0.8 by default. Run two versions on the same cases and the difference is a number.

Your own AI. It reads the saved cases over MCP, runs the prompt, scores the answers and tells you the result in the conversation. The running is done by the AI you bring.

Not in the pre-beta. You store, version and share agents and workflows, and your AI runs them. The running is done by the AI you bring.

The Model Context Protocol is the open standard AI assistants use to connect to other software. Prompt Verse offers an MCP connection, so your assistant can search, read, render, update and commit.

Add Prompt Verse as a connector in your assistant’s settings, paste your Prompt Verse address and sign in when it asks. The steps are in Use with your AI.

In the app, under Connect, in the Keys section. A key is shown once when you make it, and you can revoke it at any time. Use a key for clients that cannot sign in.

In the pre-beta you can publish a committed version, browse, try, install and withdraw. Paid checkout is mocked until payments open, so no real money moves yet.

Nothing. A listing points at a commit, never at your working copy, so what someone installed cannot change under them.

That is planned, not available. The idea is that creators sell runs on our side and buyers never see the prompt. There is no date. Ask for it as a feature if it matters to you.

Pro is €19 per seat per month, Premium €39, Max €79. Payments are in test mode during the pre-beta. See pricing.

No free plan and no trial period. We can give people we invite a promo key for a 7, 14 or 30 day trial.

Your own AI runs your prompts on your own subscription, so there is nothing to meter here. Bring your own AI on every plan.

Pro: 50 prompts, 25 workflows and 25 agents. Premium: 100, 50 and 50. Max: no counts. Versions, evals, sharing and the marketplace are on every plan.

We store them to run the service and show them to the people you choose. Your prompts reach a model only when you connect your own AI. The privacy statement is a draft until a lawyer has reviewed it.

Yes. The app has a data export and an account deletion with a two-week grace period.