/work
// the things I've built
// start here
If you only look at three: a coding-agent plugin in real use, a platform delivery that holds up in production, and the open methodology behind it.
Coding-agent plugin for any model: Claude, GPT, and more. Go CLI plus skills for memory, validation, tracked work, and reusable operating loops.
A self-hosted AI application platform adapted for restricted OpenShift with observability, repeatable deployment, and operator-safe handoff.
Operational principles for making AI agent work reviewable, testable, and safe to ship.
// the full set
Grouped by what each one proves: reliability, operations, and what is open to the public.
Tools that keep AI agents on track: they hold context, check their own work, and turn messy sessions into something a team can trust.
Coding-agent plugin for any model: Claude, GPT, and more. Go CLI plus skills for memory, validation, tracked work, and reusable operating loops.
Operational principles for making AI agent work reviewable, testable, and safe to ship.
A validation CLI that reads git history and turns AI-assisted coding sessions into measurable reliability signals.
The infrastructure that hosts AI applications where uptime, security, and recovery matter, including locked-down environments most platforms cannot run in.
A self-hosted AI application platform adapted for restricted OpenShift with observability, repeatable deployment, and operator-safe handoff.
Enterprise container-platform patterns for restricted self-managed environments: multi-tenant, GPU-aware, GitOps-driven.
Automation for upgrades, diagnostics, secrets, backups, recovery, and day-2 operations where reliability has to be inspectable.
Public work and exploration: the code is on GitHub, the thinking is in the writing.
This page is the what. The method is the how.