What is the JevOps control plane?

An AI decision reliability control plane from Trevorton27: it evaluates proposed agent actions with TypeSafe Jev for semantic judgment, applies deterministic Python policy rules, and returns a disposition, with a console for reviewing decisions, managing policies and monitoring agents.

What dispositions can JevOps return?

Four: ALLOW, RETRY, HUMAN_REVIEW and BLOCK. A failure of the model call never becomes allow; it falls back to human_review, and the error is recorded in the policy trace.

How does JevOps talk to the Jev model?

Through the typesafe-sdk package. One AsyncTypeSafeClient.system_one call per evaluation sends the action, objective, state and evidence and asks typed questions: Noul for 0-1 appropriateness, Score for risk 1-10 and Choice for the recommended route.

What happens when a Jev call fails in JevOps?

It falls back to human_review and records the error in policy_trace; the README states a failure never becomes allow. The failure provider mode (JEVOPS_JEV_PROVIDER=failure) simulates timeouts and errors for testing.

How is JevOps deployed?

The API and Celery worker run on Railway (the API's root directory must be /apps/api with healthcheck /health), Postgres on Neon, Redis on Upstash, and the console on Vercel; a make up, install, migrate, dev sequence gets it running locally.

How does Support Buddy X9000 use JevOps?

It uses JevOps as a gate on AI-drafted customer replies: POST /v1/decisions/evaluate gets a disposition, with a block failing its guardrails, and outcomes are posted back as ground truth for calibration. It treats JevOps as optional and continues with local guardrails on any error.

Does JevOps have a licence?

No licence is shown in the repository listing, so reuse terms are unclear; the project also ships a Python SDK in packages/python-sdk/ and the Jev integration is one of three provider modes.

Jev directory

CommunityJev Tools & CLIs

JevOps — Trevorton27

JevOps is a control plane that has TypeSafe's Jev judge each proposed AI-agent action, then applies deterministic Python policy rules to return ALLOW, RETRY, HUMAN_REVIEW or BLOCK, with a FastAPI API, Celery worker and Next.js console.

See highlights Open github.com

Stars: 0
Language: Python
Reading time: 2 min
Published by: Community
Author: Trevorton27
Last verified: Oct 2026

Trevorton27/jev-ops

User repository on GitHub

View on GitHub

JevOps is a control plane for AI agent decisions. TypeSafe AI's Jev model judges each proposed action, then deterministic policy rules return ALLOW, RETRY, HUMAN_REVIEW or BLOCK. Built with FastAPI, Celery (drift detection, replay runs) and a Next.js operations console.

Forks: 0
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 JevOps — endomorphosis.

Tagscommunity control-plane ai-agents decision-gating

Works forJev for Decision gating Jev for Realtime control Jev for Routing

Used inJev for Software & DevTools Jev for Customer Support

Languages: Python
Platforms: Vercel AI Gateway Next.js PostgreSQL

Highlights

Evaluates proposed agent actions with TypeSafe Jev's typed judgments, then applies deterministic Python rules returning ALLOW, RETRY, HUMAN_REVIEW or BLOCK.
One Jev call per evaluation via AsyncTypeSafeClient.system_one, asking Noul (0-1 appropriateness), Score (risk 1-10) and Choice (recommended route).
JEVOPS_JEV_PROVIDER chooses mock (default), typesafe or failure; a failed model call falls back to human_review and never becomes allow.
FastAPI API and Celery worker on Railway, Postgres on Neon, Redis on Upstash, and a Next.js console on Vercel; startup runs alembic upgrade head then uvicorn.
Support Buddy X9000 gates AI-drafted replies on the API and records approve/reject outcomes to calibrate Jev; Devin fix PRs get merge-readiness reviews.

Watch out

The repo listing states no licence, so reuse terms are unclear. It is brand new (created 2026-10-14, zero stars), the web console bundles NEXT_PUBLIC_API_KEY into the browser so anyone who can open it can read the key, and the README warns the destructive demo reset must stay disabled in production.

Questions about this entry

Short answers drawn from the linked project or page.

From the Jev guides

Original write-ups on what this entry is for.

Jev at home: voice pipelines, Assist routing, and sentinels

Oct 16, 2026

How Home Assistant conversation agents, macOS voice pipelines, and an advisory sentinel put a fast typed decision in the loop — one call per utterance, thresholds in code, and a daily token budget.

Jev Tools & CLIs#home-assistant#voice#automation

Read article

Search visibility, audited by decisions

Oct 16, 2026

Jev-based SEO tools split the crawl from the call: deterministic crawlers count and parse, while typed judgments rank fixes, place internal links, score search results, and track AI-answer mentions — with costs and caveats attached.

Jev Tools & CLIs#seo#search-visibility#audits

Read article

Jev on the Gateways: Vercel, Workers AI, and OpenRouter

Oct 12, 2026

The request shape barely changes across surfaces; the surrounding contract does. When to call the API directly and when a gateway, a Worker binding, or a router entry earns its extra hop.

Jev Tools & CLIs#gateway#vercel#cloudflare

Read article

From the Jev AI model community

Posts from builders shipping with Jev right now.

Follow @typesafeai

JevOps — Trevorton27

Highlights

Questions about this entry

What is the JevOps control plane?

What dispositions can JevOps return?

How does JevOps talk to the Jev model?

What happens when a Jev call fails in JevOps?

How is JevOps deployed?

How does Support Buddy X9000 use JevOps?

Does JevOps have a licence?

More Jev resources like this

From the Jev guides

From the Jev AI model community