I'm Matt Wood, and this is For Your Information. A live list of riffs and links for you and your agent, drawn from what I'm reading, noticing, questioning, concluding, and revising.
"There are three phases to AI coding. The first is autocomplete. The second is AI-assisted chat panels in IDEs. We are now entering the Agent-First era." — Zach Lloyd
The Agent Stack Bet - by Addy Osmani - ElevateThe three-phase model of AI coding (autocomplete, chat, agents) sets up the 'agent stack' framing that Osmani builds on as the next evolutionary stage.
Loop Engineering - by Addy Osmani - ElevateWarp's description of agentic coding as the emerging third phase leads directly into loop engineering as the practice for designing agent-driving systems.
Supported by
ChatGPT Desktop App for Linux Now in PreviewChatGPT's Linux desktop app with Codex integration directly addresses the developer workflow lessons discussed at Warp — bringing AI assistance natively into the terminal/IDE environment rather than requiring context-switching to a browser
Saggar: Terminal Session Manager for MacSaggar's session/task status management directly addresses the developer experience lessons described in building Warp — organizing terminal sessions around project state is a core developer productivity problem
Secrets Don't Belong in ConfigLessons building developer tooling at Warp align with the article's thesis: developer experience improves when secrets management is a first-class concern separated from general configuration rather than bolted on
Omarchy: Modern Linux DistributionBoth reflect lessons from building developer-focused tools with strong opinions — Omarchy's opinionated approach mirrors Warp's philosophy of making deliberate UX tradeoffs to serve developer workflows better
How I Over-Engineered My BookBoth derive lessons from first-hand building experience; the over-engineering post's cautionary reflection on excessive complexity aligns with lessons-learned narratives about pragmatic developer decisions
Related
8 Myths of Software Development with AIACM Queue practitioner-focused systems articles share the same audience and knowledge domain as lessons learned building developer tools at companies like Warp
8 Myths of Software Development with AIACM Queue practical systems articles and lessons from building developer tools like Warp share the same practitioner-audience focus on real-world software engineering insights
Forking dotenvy into dotenv-ngBoth address developer tooling reliability: dotenv-ng fixes a silent correctness bug that wastes developer time debugging authentication failures, echoing the developer experience lessons from building Warp
TerminalWidget for Mac, iPhone, and iPadBoth address developer tooling experience; TerminalWidget and Warp's lessons both focus on making terminal/developer workflows more accessible and integrated into modern interfaces