Devs hail jev as a missing primitive: compose AI into systems, not as the system
teropa · x · 2026-09-17
Developers in the HumanLayer orbit are discussing jev, a new building block for AI systems. @dexhorthy calls it the best excuse to re-read 12 Factor Agents: tool calling can be decomposed into classify+action, and if you design AI programs as pipelines that switch between classification, data structuring, deterministic code, and small agent-shaped loops, jev is a wonderful building block. @dillonmulroy argues it resonates because it unlocks composing AI into systems and products rather than AI becoming the whole system — a missing primitive. The referenced 12-factor-agents repo (26k stars) codifies principles for production-grade LLM applications.
More from coding & agent
- Matt Pocock proposes /pr skill: every model creates garbage PR bodies today — mattpocockuk · 2026-09-17
- Codex Agents Pile Up 76GB of Debug Files, Burning Through 5x Quota Overnight — craigbalding · 2026-09-17
- Build the agent setup first, pick the model second: a Linux + Tailscale + llama.cpp stack guide — max_paperclips · 2026-09-17
- Dev claims LLM memory is basically solved using LLM plus Jev for recall — JoshuaJBouw · 2026-09-17
- AI agents can now control Google Home as MCP server opens to third-party assistants — emmanuelvivier · 2026-09-17
- Bindu Reddy Teases Upgraded Agent Integrating Jev, Launching Tomorrow — bindureddy · 2026-09-17