Posts 324 entriesPage 2 of 33

Posts

Technical Deep-Dive

Wrong-Mode Green Is Not a Gate

A freshness gate that accepts the wrong mode flag and still exits green trains people to trust a board that cannot see. Pin the argv semantics, not just the script path.

Read
Technical Deep-Dive

Temporary Is Not a Plan: Fork Discipline for an Adopted LMS

Adopting an open source LMS got us live in days. What keeps a fork from becoming a private product: a named upstream path and a written retirement condition.

Read
Technical Deep-Dive

Do Not Blindly Restart: Designing a Self-Healing Watchdog That Stays Honest

A burn-in watchdog latched on a phantom io spike. The fix: fail-closed self-healing that restarts an allowlist, latches everything else, and never trusts one sample.

Read
Technical Deep-Dive

Passing Is Not Validating: A Green Check With No Teeth

A JSON Schema validator passed all 19 checks while validating nothing. How a code review proved it, and the teeth check that fixed it.

Read
Technical Deep-Dive

A Green Recovery Drill Can Still Be Lying

A recovery drill that passes green can still be lying. Learn four false-pass traps in disaster recovery and backup restore verification.

Read
AI Engineering

Let the Model Judge. Make the Code Decide.

A production writing pipeline rebuilt around one rule: the language model owns judgment, deterministic code owns the gates and the audit trail.

Read
Technical Deep-Dive

Copying Files Is Not Installing

A marketplace install copies a plugin's files into place but never runs its installer. Here is how that bricked local mode and the dependency-free bootstrap that fixed it.

Read
Technical Deep-Dive

Producer Fallback: When Claude Hits the Weekly Limit, the Pipeline Still Ships

A weekly LLM limit is a producer outage, not a reason to miss the daily post. Grok fallback + land path kept startaitools shipping.

Read
Technical Deep-Dive

Exit 0 Is Not Success: Automation Assurance That Verifies Outcomes

A green exit code can hide a dead job. Intent-OS automation assurance treats outcome verification as the success criterion, not exit 0.

Read
Architecture

Empty Is Not Clean: Five Fail-Open Bugs in an AI Agent

One fail-open bug, five disguises: empty args, an empty log, a missing property, a sheared chain, an unset scope. When there's nothing to check, fail closed.

Read