⌘ GitHub
jev-router
Route Claude Code tasks to a suitable lower-cost model from a defined set of candidates.
Model & skill routing
THE INDEPENDENT JEV PROJECT DIRECTORY
46 JEV PROJECTS.
BIG DISRUPTION.
46 ways to put Jev to work. Discover tools that route models, trim context, check code and turn everyday decisions into software.
THE NEXT MOVE IS YOURS.J / 01
46PROJECTS.
10 CATEGORIES.
PLENTY TO BUILD.
10 CATEGORIES.
PLENTY TO BUILD.
CHOICE / SCORE / NOULEXPLORE ↗
THE PROJECT INDEX / 46
Community projects, demos & independent experiments.
⌘ GitHub
jev-agent-skill-router
Choose a skill for an agent, with explicit no-skill and review paths when the match is uncertain.
Model & skill routing
⌘ GitHub
jcm-router
A local proxy that selects a Claude model and reasoning effort for each message.
Model & skill routing
⌘ GitHub
jev-router · prismhq
A LiteLLM-based router that uses Jev to choose a model for each request.
Model & skill routing
⌘ GitHub
fx
A coding agent from Vercel Labs that uses Jev for permission-review decisions.
Agent guardrails
⌘ GitHub
jev-axi
Assess shell commands for destructive actions, data transfer, remote execution and weakened security before an agent acts.
Agent guardrails
⌘ GitHub
jev-guard
A guard layer for coding agents that evaluates prompt injection and potentially dangerous operations.
Agent guardrails
⌘ GitHub
is-malicious
Inspect dependency source and build files, then examine suspicious files and lines more closely.
Agent guardrails
⌘ GitHub
pi-heed
Check whether an agent’s next action actually follows the user’s request.
Agent guardrails
⌘ GitHub
jev-git
Scan changes for secrets and destructive commands before a Git commit or push.
Agent guardrails
⌘ GitHub
Abide
Watch coding-agent changes and flag potential violations of project rules for review.
Code review
⌘ GitHub
Clean Code Judge
Evaluate pull-request files against 31 code-quality questions, then pass findings to a writing model.
Code review
⌘ GitHub
DiffJury
Triage pull requests by risk and decide which reviewer should handle them.
Code review
⌘ GitHub
limpet
Check completion criteria before an agent announces that a task is finished.
Code review
⌘ GitHub
Hunch
Write coding rules in plain language and check them locally or on pull requests.
Code review
↗ Demo / report
Browserbase + Jev
Kyle Jeong’s browser-agent demonstration: Jev chooses an action; Stagehand carries it out.
Browser & web
⌘ GitHub
jev-ultrafast
A browser-use agent that lets Jev select the next action and uses a generative model when text is needed.
Browser & web
⌘ GitHub
Jev Browser
A browser-automation project built around Jev decisions at each step.
Browser & web
⌘ GitHub
unclutter
A browser extension that identifies distracting page elements and removes them.
Browser & web
⌘ GitHub
typesafe-adblock
A Chrome extension that turns ad detection into a series of yes-or-no judgments.
Browser & web
⌘ GitHub
public-browser
Let coding agents operate a real Chrome browser, with Jev helping choose actions.
Browser & web
⌘ GitHub
Sniff Test
Check paragraphs for vague phrasing, repetitive endings and other writing patterns using focused judgments.
Writing & content
⌘ GitHub
pagegrade
Evaluate sections of a web page for clarity, writing quality and SEO.
Writing & content
⌘ GitHub
JevSlop
Evaluate note.com articles across eight dimensions and combine them into a content-quality score.
Writing & content
⌘ GitHub
citation-verifier
Use Claude to locate cited passages and Jev to assess support, with a person making the final call.
Writing & content
↗ Demo / report
Email triage · Ryan Vogel
A demonstration of categorizing email, judging priority, detecting spam and deciding whether a reply is needed.
Workflows & data
⌘ GitHub
Notra
A marketing-analytics platform exploring Jev for brand-visibility classification.
Workflows & data
⌘ GitHub
Email intent routing
Classify incoming messages as invoices or general email and send them to the corresponding workflow.
Workflows & data
⌘ GitHub
jev-curate
Filter synthetic datasets using condition judgments and confidence, saving accepted and rejected records separately.
Workflows & data
⌘ GitHub
CV screener
A resume-screening experiment with editable criteria and reusable scores. Human review remains essential.
Workflows & data
⌘ GitHub
typeful-triage
An issue-triage dashboard covering type, severity, urgency, duplicates and next steps, with recorded human corrections.
Workflows & data
⌘ GitHub
fast-jev-compaction
Keep useful tool calls and results in a coding session while removing irrelevant context without summarizing it.
Context management
⌘ GitHub
yoshi
A proxy layer for Claude Code and Codex that evaluates which conversation history is still needed.
Context management
⌘ GitHub
wakegate
Check whether a scheduled agent wake-up is necessary before waking a larger model.
Context management
↗ Demo / report
Jev Trader · Jarrod Watts
A block-by-block trading experiment. The shared demonstration shows dry-run mode; it is not evidence of profitability.
Games & experiments
⌘ GitHub
HA-Jev
A Home Assistant integration that answers questions about home state with choices, scores and probabilities.
Games & experiments
⌘ GitHub
jevinci
An experimental painter that selects pixel colors in parallel and uses confidence to vary brush width.
Games & experiments
⌘ GitHub
pacman-jev
An open-source Pac-Man experiment where Jev chooses which direction to move next.
Games & experiments
⌘ GitHub
tetris-ai
Browser Tetris playable by a built-in planner, Jev or a chat model.
Games & experiments
⌘ GitHub
Jev Pong
A Pong experiment comparing model-controlled decisions as the ball moves.
Games & experiments
⌘ GitHub
robo-harness
A robotic-arm workbench that selects joint actions from candidates and includes a spending limit.
Games & experiments
⌘ GitHub
jevcal
Fit confidence thresholds to labeled examples and estimate fallback rates, with checks for model changes.
Evaluation
↗ Demo / report
When reranking does not help
A reported test of 33,047 entries and 164 queries where Jev-only reranking did not outperform vector retrieval.
Evaluation
⌘ GitHub
NanoJev
An independent 0.6B-parameter decision-model implementation with training code, weights and data.
Independent models
⌘ GitHub
Laya
An independent non-autoregressive decision-model project supporting three question types.
Independent models
⌘ GitHub
kev
A small independent decision-model experiment based on Qwen2.5-0.5B, designed to train and run on a MacBook.
Independent models
No projects found.
Try another search or choose a different category.
UNDERSTAND THE DECISION.
The complete field guide ↗What is Jev?
Meet TypeSafe’s decision model. Context and questions go in; typed judgments come out.
Understand the model ↗Why use it?
See where smaller, faster judgments can reduce overhead—and where a larger model still belongs.
Explore the trade-offs ↗How do I start?
Define one decision, connect the API, test real examples and keep a fallback path.
Build your first workflow ↗