Mick’s Foundrysystems / automation / local AI

Independent systems practice · Australia

Mick’s Foundry

Systems, automation and local-AI tools built with explicit evidence and human control.

View the workInspect GitHub
01 / Work

Selected systems

Three substantial projects spanning AI orchestration, governed knowledge and local creative production.

implementedChecked 9 Oct 2026

B.O.T.S. 5

A local-first AI campaign harness and native Linux desktop built around bounded execution, explicit authority, durable evidence and human acceptance.

Linux V0.2 is accepted for Mick's current use, with source published and tagged v0.2.0. Its control plane adds explicit capabilities, tools, execution, receipts and plugins. Plugins remain trusted in-process code; broader Linux compatibility and installed launcher matching remain unqualified.

  • Python
  • Linux
  • AI orchestration
  • SQLite
Read case study
implementedChecked 1 Oct 2026

Organisational Memory + OMC

Repository-centred project memory with explicit authority, provenance and deterministic retrieval.

Repository governance and deterministic OMC retrieval are implemented; broader successor cutover remains deliberately incomplete.

  • provenance
  • retrieval
  • governance
  • Git
Read case study
activeChecked 1 Oct 2026

Story Audio

A resumable local-AI production system for turning long-form writing into directed, assembled audio.

The production pipeline is active and incomplete; this is not presented as a finished audiobook system.

  • local AI
  • speech synthesis
  • production planning
  • resumable pipelines
Read case study

02 / Practice

Practical work, not a title exercise.

Mick is a career changer with a background in hands-on industrial work. The current practice applies that same fault-finding instinct to Linux systems, automation, local AI and evidence-led project delivery.

He identifies the problem, designs constraints and operating procedures, directs AI-assisted implementation, and decides what evidence is sufficient for acceptance. The contribution is described accurately; neither hidden behind “the AI did it” nor inflated into sole authorship of every line.

Read the working approach

03 / Now

Current technical focus

Static, dated—not a fake operations console.

  • Bounded AI workers and acceptance workflows
  • Repository-centred organisational memory
  • Resumable local-AI audio production
  • Native Linux desktop reliability
  • Storage, recovery and operational automation

04 / Evidence

Claims should have somewhere to land.

Case studies distinguish implemented work from active or incomplete work, name the public evidence boundary, and link to source material when that material can safely be public.

All case studiesThe supporting lab