Supervise the agent, don’t worship it
The Guild focuses on the habits that keep AI-assisted work reviewable: small diffs, clear constraints, explicit tests, and stop points before deployment.
Tom Hundley has spent 25 years in software architecture, security reviews, production debugging, and system design. AI Coding Guild exists to turn that judgment into something non-software developers can actually use when AI agents start generating code.
The goal is not to make everyone a senior engineer overnight. The goal is to teach builders how to slow the agent down, ask better questions, spot obvious failure modes, and ship software that does not collapse the first time reality touches it.
Production-safe articles written for people who need better software judgment, not more hype.
OpenPublic copy-paste prompts that keep AI coding work reviewable, scoped, and testable.
OpenThe structured learning path for non-software developers who want to supervise AI agents well.
OpenThe Guild focuses on the habits that keep AI-assisted work reviewable: small diffs, clear constraints, explicit tests, and stop points before deployment.
Security boundaries, auth mistakes, secrets handling, database changes, rollout risk, and failure recovery are all translated into systems you can apply without a CS degree.