Guide
Host selection, workspace setup, memory, identity, tooling, GitHub, deployment workflows, cron/heartbeat routines, security boundaries, and operator runbooks.
Operational AI Starter System
A practical guide, starter kit, and operating framework for turning an AI assistant into a controlled engineering operator that can inspect files, make changes, verify results, and report evidence.
Full guide and starter kit remain private until access and release strategy are approved.
Most AI assistant setups fail because they optimize for conversation instead of operations. This system is designed around evidence, registries, QA checks, and human authority.
Host selection, workspace setup, memory, identity, tooling, GitHub, deployment workflows, cron/heartbeat routines, security boundaries, and operator runbooks.
Copy-ready workspace templates, project registries, release checklists, task prompts, QA scripts, monitoring examples, and placeholder-only environment files.
A public-safe Compass Infrastructure case study showing how IT Daily, PortLab, and the agent operating model became one repeatable system.
Generic AI assistants often lack durable project state, deployment records, approval boundaries, QA evidence, and scheduled follow-through. The result is plausible responses instead of reliable engineering output.
Best fit: homelab builders, IT students, junior sysadmins, solo technical operators, and small business technologists who already understand terminals, Git, and basic hosting.
This is not a replacement for human review. The system is intentionally constrained: destructive, public, external, paid, and credential-sensitive actions require human approval.
The first release path is a controlled early-access test, not a public dump of the full private repository. If you want the guide, starter kit, or setup help, contact Compass Infrastructure and describe the system you want to build.