After this chapter you will be able to use autonomous coding agents — Codex, Claude Code, Claude Cowork — to build, review, and iterate on real software: hand off whole tasks, orchestrate them from GUIs and task systems, verify their output, and know where they still fall short. The material here is drawn from practitioners who run agents on production code every day, and the recurring lesson across all of it is the same: the human's job shifts from typing code to planning, orchestrating, and verifying. Explore each theme below and understand how the pieces fit together into a working practice. [^7d2a9b32-b281-5828-9eb1-bf812129ae58@426] [^11bdf6ed-1223-5cba-afde-fa9b4ce88712@284] [^ceebdf5b-d543-5073-becc-f4bfbb1dcabf@382] [^717e6970-382c-5199-9605-8fa2c37ec1b2@102] [^619ac4a8-531a-53a9-9188-912ae7459502@157] [^f3e3b544-e101-5589-b258-66e9ef89e7b0@145] [^9ad78550-8aa6-5f8a-8069-903f60c1df13@235] [^5525946f-790a-5dbf-869f-eaf4933bd18b@138] [^8503b75a-1b68-5b82-ad35-4b11e34181c7@20] [^338007c4-7551-57c5-9af0-2811f0432ddb@132] [^96100b6a-a941-5a32-adef-3df8a7183d9e@47] [^320b90ca-d366-5cb1-9dbb-89b66cd5893b@256]