Unified Brain

unified-brain

Persistent memory for System One coding agents, by DevLoops.

On 24/10/2026 I have published as OSS a framework I call unified-brain: evidence gathering that is fast and cheap, and as accurate as a frontier coding agent. On the same nine-question, 74-point xz benchmark (evals/questions/xz.md) it scores 93.2 % — the same as a Claude Fable 5.1 agent, against 94.6 % for a Claude Opus 5.5 agent — at 1/8 and 1/20 of their cost (prompt caching on), 2.4–3.5× faster. It started at 86.5 % on 29/10; the four retrieval changes that closed the gap are each written up in evals/results/. The gathering is done by Jev, a frontier decision judge model, not by the frontier model.

The magic sauce is an intermediary OS (SQLITE) memory layer that facilitates retrieval, filtering and judging through a System One model, namely Jev.

The agent spawned from unified-brain runs through MCP on Cursor, OpenCode or Claude Code etc. allowing you to use it through your favorite harness.
So to be clear here: unified-brain is not a harness but a perception layer any harness plugs in!

The evals on two large OS libraries (xz and open62541, in evals/ on this repo) are close enough to the frontier agents, at a fraction of their cost, that I believe this is a missing link to make AI more reliable.

This is because the limits holding back AI are:

- The Reasoning and Hallucination Gap:

The LLM turn reasoning it is introducing uncertainty and probabilistic outcomes.

BUT.. Unified-brain's System One discovery engine breaks the task into hundreds of yes-or-no calls, so the search itself does not reason in prose and cannot hallucinate a path. Across five graded xz runs the answers contained one wrong claim per run at most, always the same one on the longest answer — a constant misquoted by the…