Using Jev as cheap verifiers to keep long-horizon agents on track, every turn
omarsar0 · x · 2026-09-20
Elvis Saravia (omarsar0) shares what he calls one of the craziest use cases for Jev: building a custom verifier for the /goal feature in his agent harness. The verifier checks whether the goal is actually complete after every turn, making continuous verification cheap enough to scale.
- These checks were previously handled by an expensive reasoning model; with Jev they can run more frequently, keeping agents on track
- His thesis: System One models are perfect for verification, and cleverly combining System One and System Two models extends harnesses further
- He believes this pattern will enable a new wave of scalable test-time compute methods and calls it an "insane unlock" for long-horizon agents
- Still early: he's seeing good results but needs to benchmark it; a full guide is coming in the next couple of days
More from coding & agent
- 70 hands-on cybersecurity projects with full source code — tom_doerr · 2026-09-21
- Dev compares coding models building a coop game: V4.1 outshines Astra's 'pathetic' default taste — teortaxesTex · 2026-09-21
- 'Just 3 lines of code' backfires: dev argues tools should expose complexity, not hide it — willcb · 2026-09-21
- TypeSafe's Jev returns typed decisions with probabilities, not text — here's where it fits in agent loops — prakersh · 2026-09-21
- Jev Engineering gives agents a decision brain, 193x faster and 444x cheaper in tests — agihouse_org · 2026-09-21
- Dev Argues PAW Shouldn't Hide Its Complexity, Points to DSPy as the Better Playbook — willcb · 2026-09-21