Jevonian
Local OpenAI-compatible proxy that sits between a coding agent and its providers and asks Jev, per turn, which routed model and thinking level should serve it.
One local endpoint. The right model for every turn — enforced in code, not prompts.
How it uses Jev
1. Code narrows the candidates. Configuration, wire protocol, quota health, context window, and thinking-level floor are arithmetic, applied before any model judgement. 2. Jev picks the route for jevonian/auto — model and thinking depth in a single call. 3. The proxy forwards the request and records usage, reason, and result in the ledger.
Getting started
Or with pnpm: pnpm add --global jevonian .
- Languages
- TypeScript 98% · Shell 1% · JavaScript 1% · CSS 1%
- License
- AGPL-3.0
- Latest release
- v0.1.7
- Last activity
- Sep 2026
- Added
- 2026-09-21
- Stars at snapshot
- 0
- Status
- Community listing
Source: community catalog. Details and metrics reflect the published community snapshot and are not a performance endorsement. Documentation details extracted from GitHub · the README.