Jev in production › Agent context and memory

unified-brain

Open-source coding agent that answers codebase questions using tree-sitter, SQLite memory and Jev judgments. Up to 1/18 of frontier model cost with equal or better accuracy (vendor claim).

Open on GitHub ↗

1/18measured against a baseline, as published by the source
Use
Agent context and memory
Industry
Developer tools
Form
Open-source tool
Stage
In production
Listed
2026-10-27
Found via
discord
Repository
DevLoops-ai/unified-brain
Stars
23
Forks
8
Last push
2026-10-29
Language
Python
License
AGPL-3.0

The README opens with

On 24/10/2026 I have published as OSS a framework I call unified-brain: evidence gathering that is fast and cheap, and close to a frontier coding agent in accuracy. On the same nine-question, 74-point xz benchmark (evals/xz.md) it scored 86.5 % against 94.6 % for a Claude Opus 5.5 agent and 93.2 % for a Claude Fable 5.1 agent, at 1/10 and 1/26 of their cost (prompt caching on), 2–3× faster, with zero wrong claims. The gathering is done by Jev, a small judge model, not by the frontier model.

Badge

For the project's own README, linking back here:

Listed in Jev in production

Also used for agent context and memory