Jev AgentA source-verified directory of 487 Jev projects — browser agents, model routers, SDKs and MCP servers — each one showing the exact decision it hands to Jev. Curated by the Awesome Jev community, mirrored here under MIT.
Give Jev the current state and a few options. Get the Agent’s next step.
Core call · setup in the full example
const nextStep = choice("What next?", { login: null, askUser: null, retry: null });
const { answers } = await client.systemOne({ state: "Login form is ready", questions: { nextStep } });
console.log(answers.nextStep.choice);Node.js 20+ · Set TYPESAFE_API_KEY on your server. This page shows code and makes no model requests.
A Jev guardrail example in Agentgateway using a webhook to inspect model requests and responses.
Jev scores jailbreaks, harmful content and secret disclosure; thresholds or evaluation errors reject requests.
Demonstrates gateway integration, not guaranteed prevention of all attacks.
An optional Jev command-risk advisor inside a native macOS Agent, with a TypeSafeKit client.
Adds a destructive-risk judgment after local shell checks and refuses commands above the configured threshold.
API failures generally allow the command with a warning, so it is not complete system protection.
A collection of Jev developer-tool experiments, including Commit Sentry for semantic checks on staged diff hunks.
Judges risks such as exposed credentials or destructive changes; local rules warn or block a commit.
Presents risk labels with the corresponding diff before commit.
Interlinked adds optional Jev judgments and evidence checks to local coding-agent checks.
Assesses inputs such as test names, documentation claims and supporting evidence.
Keeps model advice distinct from deterministic rules and records checks.
A browser extension using Jev to identify ads, promotions, and subscription popups and save reusable hiding rules.
Evaluates candidate page elements before the extension applies local hiding rules.
Turns page judgments into reusable rules.
A Pi extension that flags tool risks before execution and checks output for secrets and failure types.
Jev scores destructive actions, scope and output risks; configuration chooses warnings or confirmation.
Defaults to shadow mode and fails open on API errors; it is not a standalone security boundary.
Adds checks for project rules, out-of-scope actions, repeated failures, and completion claims to Pi Agents.
Uses Jev to judge rule violations, irreversible actions, and task drift, feeding results back to the Agent.
Brings rule findings and actions needing confirmation into the active workflow.
Jev commands in the WordPress toolkit CaptainCore answer structured questions and prioritize malware scanner findings for review.
Sends scan rules, matched snippets, and file context to Jev to assess likely true positives and suggested handling.
Organizes rule-based scan findings into a queue for human review.
A Discord moderation bot using Jev to check spam and scam links, with local warning and timeout rules.
Assesses messages or recent member history; the bot applies deletion, warnings or timeouts.
Connects judgments, moderation logs and manual correction controls.
Adds rule checks to Pi commands and file operations, then uses Jev to assess cases that need further judgment.
Local rules handle explicit denies and allows before Jev assesses authorization and risk for bash, write, and edit.
Keeps policy and decision records available for adjusting operation gates.
CLI security scanner using Jev to inspect source files, build scripts, and CI configs for suspicious patterns before execution.
Feeds file contents and security criteria into Jev to assess risks such as data exfiltration, obfuscated loaders, and credential leakage.
Provides rapid pre-execution safety screening across multiple file formats without needing a heavy static analysis toolchain.
Tool-call checks for coding Agents that assess operation risk, user intent and prompt-injection signals.
Jev evaluates tool inputs or outputs in session context; local rules allow, warn or block.
Adds review points around Agent tools, with behavior depending on the host client.
Jev-powered plugins for coding agents
Jev returns a structured decision for the local program; consult the source for the exact decision policy.
Adds structured choices or scores to the workflow; performance and cost benefits have not been independently verified.
An experimental Hermes command-approval provider for the auxiliary.approval task only.
Jev returns APPROVE, DENY or ESCALATE, with local thresholds and policy determining handling.
Demonstrates an approval integration; the author labels it a proof of concept, not production assurance.
Reviews Agent Skills and MCP configurations and source code with local static checks and TypeSafe Jev before installation or execution, reporting file and line evidence, risk categories, model probabilities, and coverage gaps.
Jev returns a structured decision for the local program; consult the source for the exact decision policy.
Adds structured choices or scores to the workflow; performance and cost benefits have not been independently verified.
Open-source Pi coding-agent extension that uses Jev to check tool calls before they run, scan files for prompt injection, flag risky replies, and keep secrets out of what it sends.
Before each tool call, Jev receives the user's request (or pinned task), the proposed command, the agent's explanation, and recent conversation. It answers a Choice for intent (on task / off task / injected / needs more context) and a Score for risk (safe / needs approval / harmful). Local code turns these into run, ask the user, or warn. If Jev says more context would help, the extension adds it and asks again.
Adds structured choices or scores to the workflow; performance and cost benefits have not been independently verified.
Jev is a decision model: it answers a typed question with a calibrated probability rather than prose. This action asks it one yes/no question per thing worth being sure about — answered in parallel in a single call — and approves only when every one of them clears your threshold:
Jev returns a structured decision for the local program; consult the source for the exact decision policy.
Adds structured choices or scores to the workflow; performance and cost benefits have not been independently verified.
Check Pi code edits against repository Markdown rules with TypeSafe Jev
Jev returns a structured decision for the local program; consult the source for the exact decision policy.
Adds structured choices or scores to the workflow; performance and cost benefits have not been independently verified.
An Android notification and SMS filter that applies local OTP rules before asking Jev whether a message is advertising noise.
Classifies text that passes local gates; local mapping decides whether to allow or suppress it.
Makes OTP handling and fail-open behavior explicit in code.
**Open-source AI agent monitoring, malicious-agent detection, and escalate-only containment** for sandboxed LLM agents. Local HackSpain 2026 stack (AngryRobot dashboard): FastAPI, React/Vite, Neo4j. Classifies a *chain of actions*, not a single tool call. A model never pulls the plug.
Jev returns a structured decision for the local program; consult the source for the exact decision policy.
Adds structured choices or scores to the workflow; performance and cost benefits have not been independently verified.
Adds tool-call approval and tool-list pruning to the Vercel AI SDK.
Deterministic rules run first; Jev assesses remaining call risks and tool relevance for threshold-based handling.
Approval-call failures request human review; model judgments cannot guarantee detection of every risk.
An OpenAI-compatible proxy that sits between your LLM and your users. It evaluates each sliding window of tokens **while the response is still streaming** and cuts the stream **before** a violating token can reach the screen.
Jev returns a structured decision for the local program; consult the source for the exact decision policy.
Adds structured choices or scores to the workflow; performance and cost benefits have not been independently verified.
An HTTP inspection gateway that checks request content with Jev before optionally forwarding it to an HTTPS destination.
Jev judges prompt injection or unsafe content; local rules forward, flag or block the request.
Adds an inspectable content check ahead of an existing HTTP service.
Jev (System One decision model) plugin suite for DeepSeek Harness (dsh)
Jev returns a structured decision for the local program; consult the source for the exact decision policy.
Adds structured choices or scores to the workflow; performance and cost benefits have not been independently verified.
A community-maintained directory and radar for Jev, highlighting open-source projects with verified code and clear decision architectures.
Jev is TypeSafe's decision model. This directory groups projects by how they use choices, scores and probability judgements, with implementation links to help you assess fit for your own task.
TypeSafe ↗We prioritize open-source projects with clear Jev implementation evidence. Repositories without verifiable source code are excluded from the main directory.
Directory data and design from Awesome Jev by logicrw, used under the MIT License.
