Topic filter
Operating layers for agentic AI work: roles, memory, handoffs, review, and recovery.
This tag is a topic filter. It brings together guides from different parts of the Guide library.
Browse all guidesA plain guide to changing one shared setting for AI coding tools, catching stale copies, restoring the prior value, and checking what each tool used.
Read guideA plain guide to switching AI providers without silently changing the job, dropping its controls, or hiding which route actually ran.
Read guideA plain guide to turning one recurring AI failure into a tested, reviewable rule without claiming the model learned.
Read guideA plain guide to changing AI project requirements without letting old tasks, product behavior, or checks quietly keep running.
Read guideA plain guide to building AI task and bug records that control status changes and refuse unsupported claims of done.
Read guideA plain guide to building an AI workflow that knows its current stage, blocks a skipped review, and resumes without old chat.
Read guideA plain guide for turning a messy product request into one record that AI can safely plan from.
Read guideAn autonomy matrix, exact-action contract, and reversible mutation test for AI agents that send, publish, purchase, deploy, delete, or change another system.
Read guideA practical completion-evidence contract, runnable three-case experiment, and frozen WarpOS release lesson for teams that need to verify AI agent output.
Read guideA decision matrix, bounded lane contract, and runnable collision experiment for teams using multiple AI agents in one repository.
Read guideA practical guide to recoverable AI work, with a portable resume packet, a hard-kill experiment, and evidence from a frozen WarpOS audit.
Read guideA plain-language guide to the 12 parts that keep AI-assisted work accurate, reviewable, and recoverable, backed by a frozen WarpOS audit.
Read guideA practical guide to the five records an AI workflow should keep, the smallest useful event contract, privacy rules, and a frozen WarpOS writer-reader test.
Read guideAI coding breaks when the work needs memory, handoffs, evidence, review, and recovery. That is an operating-system problem.
Read guideCLAUDE.md can give context, but serious agentic work needs hooks, validators, ledgers, and review gates that do not depend on model memory.
Read guide