answerr
Platform pairing an LLM with werr, a claimed sub-millisecond fractal decision engine that returns typed noul, choice and score answers.
Overview
answerr describes itself as a dual-cognition AI platform. An LLM side (Gemini 2.5 Flash or OpenAI-compatible) handles conversation, extracts state, writes explanations and generates code. A System One side, the author's werr engine, produces typed decisions (noul, choice, score) in under 0.5 ms with no VRAM, which the README attributes to Mandelbrot fractal escape-boundary dynamics.
Related projects extend werr to an on-chain EVM decision oracle and to a fruit-fly connectome integration. The decision vocabulary mirrors Jev's primitives, with noul expressed as ALLOWED or DENIED, but the engine is not a language model and the README's scientific claims are unusual and not independently verified.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Yes / no | Is this operation allowed? | ALLOWED · DENIED |
| Choice | Which option should be chosen?(inferred) | — |
How it works
- LLM extracts an operational state vector from conversation
- werr computes a typed decision
- Decision is returned to the app or on-chain oracle
- LLM explains or acts on the result
Strengths
- Claimed sub-millisecond, zero-VRAM decisions
Caveats
- Unconventional, unverified claims about the decision engine
- BSL 1.1 licence restricts use
- Not related to TypeSafe's Jev beyond vocabulary
At a glance
- Type
- Project
- Status
- prototype
- Runtime
- web workspace + engine library
- Has code
- yes
- License
- BSL-1.1
Use cases and integrations
How curated lists describe it
A.N.S.W.E.R.R. — The Zero-Latency Reflex AI. Bridges deliberative language models with werr's zero-memory fractal decision engine. Instant, typed decisions in <1ms.
hellogumbo/awesome-jev
Frequently asked questions
Is werr a language model?
No, the README describes a fractal-dynamics engine, not a neural model.
Are the speed claims verified?
Not independently; they come from the README.
Can I use it commercially?
It is under BSL 1.1, which limits use; check the licence.
Related projects
Oh My Pi coding agent (a Pi fork with IDE wiring) with an optional TypeSafe judgment provider for bounded decisions.
ProjectproductionRustDeepSeek Harness plugin that gives the human a Jev second opinion beside the agent's multiple-choice questions.
ProjectbetaTypeScriptDeepSeek Harness plugin registering a jev_decide tool so the agent can ask Noul, Choice or Score questions and gate on probabilities.
PluginbetaTypeScriptStandalone Node studio for building and testing workflows of typed Jev judgments with code-owned decisions, plus games and simulations.
ProjectbetaJavaScriptopencode plugin that routes the agent's small multiple-answer decisions to Jev and labels each answer auto or escalate by confidence.
PluginbetaTypeScriptSelf-hosted Yao Agents platform for agents, workspaces and task boards across devices, which announces Jev support.
Projectbeta