Read & learn
Written guides.
Understand Jev, one idea at a time. Walkthroughs, recipes, and writeups — our notes first, the original next.
Learn, build, and play with Jev — from deep technical guides to creative experiments.
Read & learn
Understand Jev, one idea at a time. Walkthroughs, recipes, and writeups — our notes first, the original next.
Watch & learn
See an idea take shape. Tutorials, demos, and deep dives, organized by topic and credited to their creators.
Make & explore
See what builders made with Jev. Games, tools, repositories, articles, and more — traced to their sources.

Every category, from games and tools to repositories and writeups, traced to a primary source.
Real-time research powered bySame ideas.
Different paths.
This is the full registry, not just the featured picks. Figures like speed, cost, and stars are maker-reported or captured snapshots, not JevMade measurements.
388 experiments · showing 61–120
Experiment with Jev from MoonBit through a typed client, command-line tools, games, policy gates, and reproducible reports.
A Pi extension that turns four semantic risk checks into allow, warn, confirm or block for each tool call.
Screen agent tool calls and returned content with Jev, then review the judgments in shadow mode before enabling enforcement.
A Claude Code experiment asks Jev eight risk questions before allowing, denying, or deferring a tool call to the built-in classifier.
Ask Jev to choose a known tool or return work to a planner, while the host supplies arguments and permissions.
Build programs and shell commands as chains of bounded Jev decisions through a TypeScript SDK and coding CLI.
A Claude Code context filter that asks Jev whether each tool result should enter the conversation.
Edit and run community-built Jev experiments for classification, comparisons, routing, extraction, policy checks, games, and simulations.
A Claude Code stop hook that checks whether a claim of finished work is backed by later passing checks.
Run an offline agent loop whose local decision model checks whether to act, continue or stop without another language-model call.
Turns subjective code-quality rules written in JSON into stable command-line and editor diagnostics.
Route Claude Code prompts, check edit rules, and gate subagents with typed Jev judgments.
A completion check for coding agents that asks for evidence of edits and tests, not just “done.”
Shorten a Claude Code session while keeping moved tool results in a drawer the assistant can read later.
A zero-dependency prose linter for common signs of AI-generated writing.
Reviews one proposed coding action with narrow Jev questions before deterministic code prepares evidence for host authorization.
patdown uses Jev to block, steer, or flag agent work against repository rules and team conventions.
A shell-friendly Jev CLI for picking, rating, checking, ranking, triaging, and guarding.
A Pi extension that asks Jev which tool fits the task at hand.
Walks an agent down a capability tree and returns only the command documentation that fits its current task.
Yoshi trims old messages from Claude Code and Codex conversations, using Jev to decide what's still relevant.
Find code by describing what it does, without first building or maintaining an embedding index.
A Pi extension asks Jev about task characteristics, then applies a deterministic Pareto policy to recommend an OpenRouter model.
A Pi extension ranks file excerpts and skill hints with Jev while caching identical judgments and preserving local fallbacks.
A Pi auto mode that uses Jev to approve shell, write, and edit calls by meaning.
Add bounded routing, risk, compaction and shell-gating decisions to an agent, while its main language model handles the writing.
Choose a model at Pi task boundaries with cached Jev classifications, conservative policy rules, and observable failover.
Try an advisory claim-checking example that keeps missing evidence and editorial approval separate from model judgments.
A Python MCP server for classifying, scoring, checking, matching, and screening with Jev.
An extension that lets DeepSeek Harness ask Jev to choose tools, check policies and select evidence.
A Pi guard that checks commands with local rules and asks Jev about the ambiguous ones.
Screens agent skills and MCP code before installation by combining static evidence with security judgments.
A memory layer that asks Jev which facts to keep, retrieve or forget, preserving the original words.
Guides agents through designing and evaluating bounded decision systems, using Jev as one hosted example.
Review pull-request diffs with typed risk questions, then turn Jev's probabilities into findings and a merge-risk verdict.
A configurable semantic linter that uses Jev to judge entire files.
Tell Claude Code when a short answer should do.
A lightweight Jev router for choosing models, tools, and subagents.
A Jev-powered scanner that checks a codebase for potentially malicious code before you run it.
A GitHub Action that approves a pull request only when every configured Jev policy question clears its threshold.
Adds Jev routing to Oh My Pi, with models, prompts and thresholds you can edit in configuration files.
Checks changed code against Markdown specifications and reports likely drift in CI or an agent workflow.
Select a coding model and reasoning effort for Claude Code or Codex, then keep that pair stable through the task.
A set of bounded Jev workflows for coding agents.
A Pi extension that routes tasks to models through Vercel AI Gateway using Jev.
An opt-in Hermes plugin that uses Jev to select one skill before the main model call.
A command-line filter that uses Jev to match lines by a natural-language description rather than a text pattern.
Turns a Git diff into native runner filters for the tests most likely to be affected.
Adds Jev-backed output pruning, injection screening, and completion gates to DeepSeek Harness.
Guides coding agents from a possible Jev use case to a bounded design, baseline, fallback, and small validation experiment.
Chooses which discovered MCP tool Codex should call, then applies deterministic policy before execution.
Defines reusable “jevels” for operational decisions and records their answers in a terminal interface.
A Python skill router that uses Jev to make typed, confidence-aware selections.
Connect Jev to an MCP client, then compare its answers with those of general-purpose models.
An agent decision toolkit with a Jev MCP server, embeddable library, and Claude Code plugin.
A Claude Code plugin for using Jev.
A fast prose linter with Ruff-style rule codes and Jev-backed semantic checks.
Jev Guard checks coding-agent tool calls and returns an allow, ask, or deny decision based on risk and session context.
A TypeScript command-line client for TypeSafe's Jev model.
Write code-review rules in everyday language, then let Jev flag the pull-request lines that appear to break them.
Jev, briefly
A chat model generates text token by token and hopes you parse it. Jev never generates a word. Your code sends the state of the world plus typed questions; Jev returns every answer in one parallel pass — typed values with calibrated probabilities and a confidence score your code can trust. About 70–500 ms end to end, $0.042 per million input tokens, output free. Trained with what TypeSafe calls Reinforcement Learning for Calibrated Decisions.
You define the options. Jev returns the chosen option, a probability for every option, and a confidence. The workhorse of routing, agents, and games — the answer space is always legal, so there is nothing to parse and nothing to hallucinate.
You describe ordered levels — say trivial / normal / critical. Jev returns the level, the probability of each, and a confidence. Scores turn fuzzy judgment (“how severe is this log line?”) into a number-free decision your code can branch on.
You assert a statement; Jev returns the probability that it is true — a noul. Moderation, verification, guardrails, “does this diff actually fix the bug?”: one question, one calibrated probability.
Official material lives at typesafe.ai and docs.typesafe.ai. JevMade is an independent community registry — not affiliated with or endorsed by TypeSafe AI.
Recurring lessons
Games, drones, trading bots, and browser agents all converge on the same loop: serialize the state, ask one decisive question, act, repeat. Jev’s latency makes the loop feel instant — the model lives inside the control loop, not outside it.
The answer says what; the confidence says whether to act. The most reliable entries threshold on confidence to route edge cases to a slower model or a human — automation with an honest escape hatch.
Makers rarely ask Jev to pick from everything. Local tactics prune 225 gomoku moves to ~40; DOM filters turn a page into an element table; code narrows options, Jev judges within them.
A recurring split: Jev makes every decision cheaply and instantly, and a small LLM is only invoked when a human-facing string must actually be written. Decision and generation are separate budgets.
Send the link — repo, live demo, post, or video — plus a line on what it does and which primitives it uses. Every entry is verified against its primary source before it ships.
Submit your experimentSign in first so we can reply to you. Submissions are reviewed, not published automatically.