Humanreview on Thomas Talks AI: AI Agent Security, Claude Code & Production Engineering https://thomasdevos.com/tags/humanreview/ Recent content in Humanreview on Thomas Talks AI: AI Agent Security, Claude Code & Production Engineering Hugo -- 0.145.0 en-us Thomas De Vos © Sat, 06 Jun 2026 10:44:54 +0100 Claude Code needs a rollback note before code https://thomasdevos.com/posts/2026-06-06.10.44.54-claude-code-needs-a-rollback-note-before-code/ Sat, 06 Jun 2026 10:44:54 +0100 https://thomasdevos.com/posts/2026-06-06.10.44.54-claude-code-needs-a-rollback-note-before-code/ A practical rollback-first workflow for Claude Code and AI coding agents: require an undo path, data-risk note, owner, trigger, evidence, and human approval before production-adjacent work starts. Claude Code human review is a control, not a vibe check https://thomasdevos.com/posts/2026-05-24.11.38.00-claude-code-human-review-is-a-control/ Sun, 24 May 2026 11:38:00 +0100 https://thomasdevos.com/posts/2026-05-24.11.38.00-claude-code-human-review-is-a-control/ A practical human-review model for Claude Code and AI coding agents: define stop conditions, review evidence, boundary pressure, rollback notes, and approval gates before production-adjacent changes merge.