Openship
This site ships its own source.
Memorioso exists because human-authored work should be verifiable rather than taken on trust. Openship applies the same idea to the software itself. Instead of pointing at a repository you would have to go and find, this site serves its own source code — every file, at the commit that is running right now — over ordinary HTTP.
There is nothing to install and nothing to sign in to. An agent that knows only the address of this site can discover the source, read how the project is structured, fetch every file, and rebuild the repository from scratch.
This build
- Commit
- ce6de3ab9bc9
- Files
- 341
- Size
- 1.62 MB
- Digest
- sha256:6df0cf9ba87cc66da6f646118e566da0386b53ac645c8ec53bde85455a165d17
Point an agent at it
Give a coding agent this address and it has everything it needs:
https://memorioso.xyz/openship/agent.txt
Or take the whole thing yourself:
curl -sL https://memorioso.xyz/openship/source.tar.gz | tar xz
What you get
A Next.js application for human-authored publications. Users sign in with World ID 4.0, write drafts, sign publication payloads through IDKit proof verification, and register the publication and its proof on World Chain through the Libro registry.
- Next.js 16 App Router (cacheComponents enabled)
- React 19
- TypeScript 5.7
- Tailwind CSS 3 with shadcn-style primitives in components/ui
- Postgres via raw SQL through pg (no ORM)
- TipTap 2 for draft and publication content
- World ID 4.0 IDKit for session proofs and publication proofs
- viem for World Chain reads and writes
- Foundry for the Libro contracts
- Vitest for tests
Openship is a protocol
None of this is specific to Memorioso. Openship is a small convention — a discovery document, a manifest, and a way to fetch files — that any site can implement so that its source is reachable by an agent that only knows its address. Memorioso is the reference implementation. The specification is OPENSHIP.md, served, of course, by the protocol it describes.
What Openship is and why it exists