io.github.lawrence3699/jev-style
Jev-Style
Local 2B and 0.8B decision models: calibrated yes/no, choice and score answers for your agent.
alive probed 2026-10-15
5tools
1.7ktokens in Claude Opus 5.5
1.7ktokens in Claude Code
0.8 %of a 200k context window
995o200k_base estimate
1.8 sto initialize
4 msto list tools
- Run it
- uvx jev-style==0.4.0 (stdio, probed in a container with the network off)
- Host
- pypi.org
- Reports itself as
- jev-style 0.4.0
- Capabilities
- prompts, resources, tools
- Instructions
- 809 characters
- Claude count
- Claude Opus 5.5, measured 2026-10-15. Counted again when the tool definitions change. Method.
- Repository
- https://github.com/lawrence3699/jev-style
- Website
- https://jevstyle.com
- Registry
- active, updated 2026-10-02 · registry entry
- Reachable
- 1 of the last 1 probe
Most expensive tools
| Tool | Tokens (o200k) | Share of this server |
|---|---|---|
| decide | 311 | 31 % |
| noul | 219 | 22 % |
| choice | 218 | 22 % |
| score | 206 | 21 % |
| model_info | 41 | 4 % |
Badge and feed
[](https://mcpcost.ulehla.dev/s/io.github.lawrence3699/jev-style)
An entry for every probe that changed the result, the tool count or the token cost: Atom feed. Add it to a feed reader or a Slack RSS app to hear when the server stops answering or its tool set grows.