What does Jev actually do that the main LLM does not?

Jev answers narrow typed questions with calibrated probabilities in a few hundred milliseconds and produces no text. The LLM writes code and plans, while Reflex code owns thresholds and escalation.

Why did a harmless command pass but git push --force ask?

There are no static allowlists: npm test passes on its own while git push --force asks, and protected paths ask regardless. You can allow once, exact action, or a session scope like all git commands.

What happens when a shell command fails or dumps huge output?

A missing dependency gets a proposed install command that is never run, transient failures get one retry, and the same failure three times tells the agent to stop and ask. Long output keeps head, tail and errors with the rest trimmed.

Do I have to enable file hints, voice and browser automation together?

No. File hints are off by default and only speak on about a third of tasks, while triage, trimming, pruning and outside-result screening can each be turned off. The browser agent stops before payments and never types passwords.

What does one Jev reflex cost, and how long does it take?

About $0.00004 per decision and 350 to 450 ms, so according to the README a whole session's reflexes usually cost less than one LLM turn.

How much does it trim from long tool output?

Measured on eight long logs, 80% fewer tokens with every needle kept. The head, the tail and anything that looks like an error, failure or diff are kept whatever Jev answers.

Where can Reflex reach Jev from?

TypeSafe's own API with a TYPESAFE_API_KEY, or OpenRouter's System One endpoint with an OPENROUTER_API_KEY. The model ids differ between providers, so Reflex remembers your model per provider.

CommunityJev Agents & MCP

Reflex — kaustav1996

Coding agent on Pi that pairs any OpenRouter LLM with TypeSafe Jev reflex checks on tool calls, turns and transcripts, where code thresholds decide allow, ask or block.

See highlights

Open github.com

Stars: 5
Language: TypeScript
License: MIT
Published by: Community
Author: kaustav1996
Last verified: Oct 2026

kaustav1996/reflex

User repository on GitHub

View on GitHub

A coding agent and personal assistant with System One reflexes (TypeSafe Jev) on top of the Pi coding agent

Forks: 1
Last pushed: Updated Oct 2026

Repo stats from the GitHub API, cached Oct 2026.

This is one of several Jev listings for the same project. The canonical entry is Reflex — kshetrajna12.

About this project

A coding agent and personal assistant that thinks with an LLM and reacts with TypeSafe. A calibrated System One model checks each tool call, turn and voice transcript in about 400 ms, and code decides what happens next.

Quoted from the project's README. Read the full README.

Tagscommunity awesomejev catalog-import

Works forJev for Classification Jev for Tool calling

Used inJev for Software & DevTools

Languages: TypeScript

Highlights

Action gate checks destructiveness, workspace scope, secrets and privilege per tool call.
Progress, completion, failure-triage and output-trimming judgments run after each turn.
Model router selects fast, default or strong tiers with per-request cost awareness.
Reflex decisions cost about $0.00004 and take 350 to 450 ms per check.

Watch out

Built on the Pi coding agent with optional voice, browser and MCP features; file hints stay off unless enabled. Copy reviewed against the repository README on 2026-10-04, not from running it.

Questions about this entry

Short answers drawn from the linked project or page.

More Jev resources like this

JevLoop

Runnable agent harness that moves pick, score and yes-no loop choices to Jev or Laya in milliseconds while reserving the language model for writing, compiling DECISION.md into executable questions.

Jev Agents & MCP#community#jevbest#catalog-import

Community

zjunlp/JevLoopTypeScript

patdown

Node CLI that lints a file tree against fuzzy natural-language rules in one Markdown file, judging each matched file with a swappable TypeSafe-backed judge.

Jev Agents & MCP#community#jevbest#catalog-import

Community

tyler-dot-earth/patdownTypeScript

jev-cli

Command-line toolkit that wraps Jev judgments into scriptable verify, screen, classify, extract, find, rerank, match, route, ask, compact, and batch commands.

Jev Agents & MCP#community#jevbest#catalog-import

Community

Nasrallah-AL/jev-cliTypeScript

jevwire

Wires Jev into agents as seven MCP tools, an embeddable JevDecisionModel library with run functions, and Claude Code hooks at tool and turn boundaries.

Jev Agents & MCP#community#jevbest#catalog-import

Community

Brainwires/jevwireTypeScript

jev-recipes

TypeScript recipe library with 248 focused functions that call Jev for routing, reranking, gating and labelling, returning structured ready-or-review decisions for Node backends.

Jev Agents & MCP#community#awesomejev#catalog-import

Community