Organization repository on GitHub
Kilo Cloud monorepo
- Forks
- 19
- Last pushed
- Updated Oct 2026
Repo stats from the GitHub API, cached Oct 2026.
CommunityJev Integrations & Adapters
Kilo Code Auto-Routing
Merged PR in the Kilo Code cloud monorepo that replaces its LLM auto-routing classifier with one OpenRouter /systemone Choice request to typesafe/jev-1.13; the PR reports 0.93-0.94 route accuracy on 72 golden cases at roughly $0.00028 per call, with a 250 ms transport-error retry.
- Stars
- 15
- Language
- TypeScript
- Published by
- Community
- Last verified
- Oct 2026
About this project
Monorepo for the Kilo Code cloud platform: the web app, Cloudflare Worker services, and shared packages. See DEVELOPMENT.md for the full local setup guide — prerequisites, database, environment variables, and common commands. For shared web env var additions or rotations, use pnpm web:env set rather than editing Vercel projects, tracked dotenv defaults, or 1Password separately. Quick start (assuming prerequisites are installed):
Quoted from the project's README. Read the full README.
Used inJev for Software & DevTools
- Languages
- TypeScript
- Capabilities
- Requires API key
Highlights
- The System One classifier reads only taskType and subtaskType from one Choice question; the other axes go to analytics only.
- Verified live on 72 golden cases through OpenRouter /systemone with no errors, route accuracy 0.931-0.944, and p50 latency around 530 ms.
- The PR removed the LLM prompt, JSON repair parser, invalid-output retry, heuristic fallback, and the @openrouter/sdk dependency from three packages.
- The benchmark winner record now carries engine: 'system-one'; if the worker deploys before the benchmark, it rejects the old cached chat-model winner.
- One retry after 250 ms covers transport errors, 429, and 5xx; other failures raise ClassifierRunError with a stage label.
Watch out
This entry tracks one merged PR (#7223) inside the Kilo Cloud monorepo, whose license GitHub reports as NOASSERTION; server-side PUT routes still accept any model id while admin pickers allow only System One ids.
Questions about this entry
Short answers drawn from the linked project or page.
It replaced the LLM classifier and its prompt, JSON repair parser, retry-on-invalid-output, and heuristic fallback with one System One Choice request to typesafe/jev-1.13 through OpenRouter, removing the @openrouter/sdk dependency from three packages.
The PR reports route accuracy of 0.931 to 0.944 on 72 golden classifier cases, versus 0.785 for the replaced google/gemini-2.5-flash-lite classifier, with p50 latency around 530 ms and roughly $0.00028 per call.
It reads only taskType and subtaskType from one Choice question sent to typesafe/jev-1.13 through OpenRouter; the other routing axes go to analytics only.
One retry after 250 ms covers transport errors, 429 and 5xx responses. Any other failure raises a ClassifierRunError carrying a stage label instead of silently falling back to a heuristic.
It removed the LLM prompt, the JSON repair parser, the retry-on-invalid-output path, the heuristic fallback and the @openrouter/sdk dependency from three packages. One System One request replaced all of that machinery.
The benchmark winner record now carries engine: 'system-one'. If the worker deploys before the benchmark is regenerated, it rejects the old cached chat-model winner rather than serving it.
The entry tracks one merged PR (#7223) inside the Kilo Cloud monorepo, whose license GitHub reports as NOASSERTION. Server-side PUT routes still accept any model id while admin pickers allow only System One ids.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
Facebook: Intake Router keeps diagnosis out of the answer space at the front desk.
FacebookClinicnotes
YouTube: Jev explained for routing and judging, with a comparison against LLMs.
Hacker News: Vercel on using Jev for defined decisions and GPT-6 Astra when the job also has to write or use tools.
flashbrew4 pointsTypeSafe AI Jev vs. GPT-6 AstraRead the thread on Hacker NewsHacker News: ASTGL replay chose local Laya routing over hosted Jev on one Mac Studio setup.
Jmeg8r15 pointsDoes an open-weight decision model beat a hosted one? Jev vs. LayaRead the thread on Hacker NewsMore Jev resources like this
12.4kGitHub stars
Open-source AI email assistant (Next.js, Postgres, 12k+ stars) with a decision-model layer that can run on openrouter:typesafe/jev-1.13: Jev-backed questions cover rule selection, cold-email detection, thread status, recurring patterns, unsubscribe pages, and sender categorization, each with evals.
Jev Integrations & Adapters#community#email#gmail
Community
elie222/inbox-zeroTypeScript
elie222/inbox-zeroTypeScript0GitHub stars
A Pi coding-agent extension that uses Jev to classify each request and route it to one of three configured models. Supports TypeSafe, Cloudflare AI Gateway, Vercel AI Gateway and OpenRouter backends, with doctor checks, shadow mode and an off-by-default setup.
Jev Tools & CLIs#community#coding-agents#routing
689GitHub stars
The Vercel AI SDK provider for OpenRouter, including openrouter.decisionModel for typesafe/jev-1.13 evaluations with providerMetadata answers.
Jev Integrations & Adapters#community#typescript#openrouter
73GitHub stars
High-performance Go event bus with a bounded MPMC ring buffer, fan-out and ordered handlers, plus an optional routing layer that evaluates deterministic rules first, reuses cached decisions, and asks Jev a typed Choice only for ambiguous cache misses.
Jev Integrations & Adapters#community#go#event-bus
Community
Protocol-Lattice/GoEventBusGo
Protocol-Lattice/GoEventBusGo0GitHub stars
A Next.js dashboard and daily Vercel cron that scores public job postings against one resume with Jev through the Vercel AI Gateway, then applies through a Grok Bot webhook that fills ATS forms and writes status back.
Jev Integrations & Adapters#community#job-search#vercel
Community
mcgalleg/grokbot-jev-jobsTypeScript
mcgalleg/grokbot-jev-jobsTypeScript856GitHub stars
Verifiable coding agent runtime with 855 stars. Ships a first-class Jev structured-output provider so agent decisions come back typed through the same decision resolver as its other providers, alongside plain-English workflows, approval gates and evidence-backed verification stages.
Jev Integrations & Adapters#community#coding-agents#verification
Community
bastani-inc/atomicTypeScript
bastani-inc/atomicTypeScriptFrom the Jev community
Posts from builders shipping with Jev right now.