Ratchet Agentic Engineering. Prototypes and tailored solutions.

Ratchet Agentic Engineering: software built by AI agents inside quality rules that can only get stricter, with a person deciding every release. Tailored applications and platforms, from a written brief to production.

• 5.942 tests on every release
• 235 / 427 requirements covered by a test
• 2.0.10 Seantral in production
• 8 articles, post-mortems included

For anyone with an idea

An application shaped on your own work, to try while it takes shape.

Who it is for. Anyone who wants to try an app or a platform built around their own workflow.

What you get. A brief generated with AI in a few minutes, a prototype that runs, a platform that grows with tests.

How to start. Describe the idea in a guided chat; the brief is generated even without an account.

1. Describe the idea: a few minutes, in a guided chat
2. Brief: confirmed requirements and a roadmap
3. Prototype: it really runs, try it on a phone
4. Platform: grows with tests, one release at a time

Describe the idea

Compared with no-code platforms

Two ways to get to an app. The full comparison also says when no-code is the right choice.

• Prompt-to-app: a prompt, an app generated and hosted by the platform
• RiftSeed: a written brief, agents building inside tests, a person who checks, a standard stack you own

The comparison

The method

Agents write the code from a written spec, inside tests that only ever tighten; a person checks the behaviour of every build and decides what goes live. The method is called Ratchet Agentic Engineering: numbered requirements, every requirement proven by a test, impact analysis before every change, independent verification, an authority that signs the release. Six practices, each with an excerpt of the file that applies it:

1. Spec before agents
2. Ceilings that only go down
3. Cross-contracts
4. Probes on the published site
5. Annotation with date and verdict
6. Gate before release

The proof: Seantral, in production

Seantral: Sea conditions for 300 Italian locations, hour by hour, with a score for eight activities. In production, open to everyone, no account needed. The dossier tells how it was built; the public post-mortems tell what broke and which test now prevents it:

• Post-mortem: eleven hours of Seantral on the wrong backend
• Post-mortem: the AI spending cap and the fallback that said nothing

Explore

• Try the apps: Seantral is open to everyone, the other prototypes open with a free account.
• The blog: 8 articles on how each app was built, post-mortems included.
• The lab: the control loop, the memory carried between projects, what stays human.
• RiftSeed and no-code platforms: two flows, four axes, and when no-code is the better choice.
• Describe the idea: the assistant turns it into a brief with requirements and a roadmap.
• Contact: the chat on the site or [email protected].