Jev directory
CommunityJev Agents & MCP
JEV Guard — AlexBzal
Open, configurable guard for company AI agents built on Jev. Agents call it before sending email, launching campaigns, or editing a CRM, and receive allow, block, or hold. Companies are described in a YAML profile; includes a demo UI, human review inbox, and eval scripts.
- Stars
- 0
- Reading time
- 2 min
- Published by
- Community
- Added
- Oct 2026
This is one of several Jev listings for the same project. The canonical entry is Jev Guard — leepokai.
Highlights
- Agents call POST /v1/guard/check and obey an allow, block, or hold verdict.
- Runs without keys: without TYPESAFE_API_KEY it uses a local Jev simulator.
- Companies are described in a YAML profile with brand, offer, policies, agents, and permissions.
- Requires Python 3.10+; the web UI is plain HTML and JavaScript with no Node build.
- Ships eval, calibration, and benchmark scripts plus a human review inbox.
Questions about this entry
Short answers drawn from the linked project or page.
No. Without TYPESAFE_API_KEY it uses a local Jev simulator and the interface indicates that. Setting TYPESAFE_API_KEY switches the guardian to real Jev.
The agent posts to /v1/guard/check with fields such as agent_id, action_type, task, and the action payload, then obeys the allow, block, or hold verdict. A Python guarded decorator wraps the same call.
Code rules handle exact cases, Jev makes the fast judgment in milliseconds, and doubtful cases go to an LLM or to a person for review.
Four screens: En vivo for simulated agents with live decisions, latency, and cost; Laboratorio for editing an action and seeing the verdict; Bandeja for held actions to approve or reject; and Metricas for quality against labels, decision routes, latency, and cost.
The service sleeps after 15 minutes without visits and takes about a minute to wake, the decision log does not persist across restarts, and the demo controls are public, so anyone with the link can start the simulated agents or reset the log.
No. All data in the included profiles is fictional, using invented people and .example domains, and the 30X hackathon case is kept only as an example profile.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
YouTube: Kubernetes admission control with a decision model, fail-closed.
Hacker News: SafeDep Gryph plus Jev checks for compromised coding-agent behavior.
2 pointsDetecting Compromised AI Coding Agents with Jev and GryphRead the thread on Hacker NewsX: Jev and Laya on security queries
X: Testing Jev for prompt injection
More Jev resources like this
GitHub issue triage agent built with Eve and Jev. It takes one typed decision per issue through a pipeline of classify, validate, duplicate and apply steps, with dry-run defaults, a label allow-list, and raw Jev answers stored for threshold tuning.
Jev Agents & MCP#agents#github#madewithjev
Communitybenjamincanac/tia
Early-stage macOS app and Chrome extension that steers X's For You feed by browsing, reading, and interacting with posts. Posts without images are judged by Jev; posts with attached images go to Luna. Reading history stays on the local machine.
Jev Agents & MCP#madewithjev#agents#browser
CommunitySkyWuu/x-feed-pilot
An MCP proxy and hooks adapter that screens an agent's tool calls before they run and the tool results before the agent reads them, with thresholds in a policy file.
Jev Agents & MCP#community#mcp#security
A Hermes Agent plugin that replaces the auxiliary approval reviewer with TypeSafe's Jev decision model: one request asks six typed questions and deterministic code returns APPROVE, DENY, or ESCALATE.
Jev Tools & CLIs#community#python#agents
Community
anpicasso/hermes-jev-approvalsPython
anpicasso/hermes-jev-approvalsPythonAn agent harness where Pi runs the loop, Gemini does the work, and Jev makes the small decisions: which model tier to use, which runbook sections to load, whether a tool call may run on a real Postgres database, and whether the report holds up.
Jev Agents & MCP#community#python#agents
Community
PromtEngineer/jev-harnessPython
PromtEngineer/jev-harnessPythonA research runtime for bounded decision models: it pages and refines a virtual option space so Jev can choose among far more tools, memories, and context blocks than a single request allows, with explicit approval for side effects.
Jev Agents & MCP#community#python#agents
From the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.