Swap in decision model Jev to end agent loops without extra LLM calls
_nerdai_ · x · 2026-10-01
The author added a bonus example to Chapter 4 of Build a Multi-Agent System (From Scratch): after each step, the agent must decide whether to continue or return the result. Originally the backbone LLM made this call via structured output—and on the final step it also wrote an answer that got thrown away.
The new example hands the decision to @typesafeai's Jev model: it receives the execution state (instruction, steps so far, latest response) and answers one yes-or-no question via the Noul question primitive, returning a probability score. If it clears a threshold, the loop returns the result with no LLM call at all; otherwise the LLM writes only the next step.
More from coding & agent
- AI Agent Polled Every 15 Minutes and Snagged a Fully-Booked 6-Seat Tokyo Restaurant in 13 Hours — armand_ruiz · 2026-10-01
- Higgsfield launches ChatGPT extension bringing computer use into Codex — SimplyAnnisa · 2026-10-01
- LibraryDesignBench tests whether AI agents can design and effectively use code libraries — a1zhang · 2026-10-01
- Ace launches as a digital employee with its own computer, email and logins — garrytan · 2026-10-01
- Anthropic's cost-optimize skill audits Claude API spend per completed task — udmrzn · 2026-10-01
- Dev rant: CI vs GitHub — the fierce fight over who slows you down — garrytan · 2026-10-01