A browser agent that finds flights in seven seconds. A Claude session cut from a million tokens to 86,000 in one second. A WHERE clause that reads plain English. Vercel's safety reviewer. We are collecting what people built with Jev in its first days, and what they argued about, each in their own words. Nobody in these bylines is on eChai, and every entry links back to where it was posted.
7 entries Clear
Alex Volkov, of Weights & Biases, ran Tamara Tran's plugin on a real Claude session: nearly a million tokens down to 86,000 in one second.
Alex Volkov
22h
This is actually insane. This uses @typesafeai Jev model, as a plugin in Claude to review all the un-nesseasary tool calls, and it takes 1s to run!
Like, literally, 1 second to take my Claude session from nearly 1M to ... 86K tokens! 😮
Ask your claude to install it and be amazed
Use this prompt
```
Install, and configure :
https://github.com/tamaratran/fast-jev-compaction
```
Tamara Tran's is the post the week's biggest idea came from. Every coding agent compacts a full context window by asking a model to summarise it; her plugin instead has Jev score every tool call in the transcript and drop the ones that no longer matter, in about a second. It is open source as fast-jev-compaction, and Diogo Almeida's reply was that it frees coding agents from designing around the KV cache.
Tamara Tran
1d
found the perfect use case for @typesafeai Jev:
instant compaction
in 2026, why is compaction still a summarization prompt?
Jev can make it instant by scoring every tool call and dropping what’s irrelevant
Riley Brown, building an agent whose choice of model is made by Jev per request.
Riley Brown
1d
Building an agent with model router powered by Jev.
Milind S on OpenMausBot: Jev reads the incoming task, decides which agents to wake and which model each one gets.
Milind S
1d
got @typesafeai's new model Jev as a chief of staff for bots
Jev reads the task, wakes the right teammates off the bench
and gives each one the right model
It is possible on OpenMausBot as it supports all the LLMs from your existing subscriptions
Jev as a decision engine is great
Niaz Morshed's jev-review is an MCP plugin a coding agent calls while it works: it scores the work on a set of metrics, the agent improves it, and the loop runs again.
Niaz Morshed
1d
built `jev-review` @typesafeai
it's an experimental, local-first MCP plugin that gives coding agents a score quality feedback loop across different metrics.
agents call jev while they work, get scored, make improvements, and repeat the loop
try below 👇
Joey Kudish's first Jev package, an MCP server so an existing agent harness can call Jev to verify a claim against evidence, screen content before it enters the context, or rank candidates by meaning. Open source at the link.
Joey Kudish
1d
Built a proof of concept MCP for @typesafeai 's new jev model
Use it inside your existing agent harness to verify claims against evidence, screen content before it enters context, and rank candidates by meaning.
Lahfir runs Claude Code, Codex and opencode side by side and had been routing between them with rules in a system prompt. A Claude Code hook now asks Jev before delegating: mechanical tasks to Haiku, harder ones to Opus subagents, long implementation work out to another harness.
Lahfir
2d
Jev solved local harness/model routing
I use a combination of Claude Code, Codex and Opencode as my local agentic stack and routing to other harnesses was always enforced in the system prompt/rules
With a deterministic hook that Claude Code can decide before delegation, Jev helps to route to the right harness/model based on the task, and it's pretty accurate based on the intensity/intelligence of the task
> Mechanical tasks get routed to Haiku
> Intelligent ones to Opus sub-agents
> Long-running implementation work to external harness
