P suite
Summary
P suite brings together my AI-coded apps: projects whose code is written almost
entirely by LLM coding agents, under my direction. Each starts with a P and
solves a different problem.
The apps live in their own repositories and share no application code. What
connects them is a common way of working, from the first development command to
a signed release.
The apps
pipelines, clusters and AI agents into one window.
screen to a build on a breadboard.
and workspaces.
The collection grows one app at a time.
A shared way of building
The P suite repository holds the conventions, starter templates and audit tool
used across the apps.
• A common set of commands for setup, development, formatting, checks,
changelogs and releases.
• Local checks before commits, covering formatting, linting and tests.
• Signed release files with checksums and minisign signatures.
• Written architecture decisions and instructions that help coding agents work
consistently across repositories.
• An audit script that checks each app against the suite's conventions.
The apps and the shared tooling are free and open source, under MIT OR
Apache-2.0.
AI-coded, human-directed
The coding agents write almost all of the application code. I choose what to
build, direct the work, test the apps and decide what ships. P suite brings
these projects together and makes that development approach explicit.