ArkEnv

ArkEnv is a typesafe environment variable validation library for modern JavaScript and TypeScript frameworks. It provides a single canonical surface across all frameworks: import { env } from "./env".

In v1, ArkEnv offers two first-class validation engines:

• @arkenv/core: Built-in ArkType DSL engine for high-performance schema definition and automatic coercion.
• @arkenv/standard: Standard Schema engine supporting any compliant validator (Zod, Valibot, etc.).

---

Capabilities

Core usage

• Define typesafe schemas using @arkenv/core ArkType DSL strings or @arkenv/standard with Zod / Valibot.
• Automatic coercion (string.numeric, number.port, booleans, JSON objects, arrays) and default values ("boolean = true").
• Single, uniform developer experience: import env directly across frontend and backend code.

Framework integration

• Vite (@arkenv/vite-plugin): Plugin operates in transform mode, rewriting imports from env.ts in the client graph to inline public VITE_* literals and replacing private server keys with throwing runtime guards.
• Bun (@arkenv/bun-plugin): Plugin intercepts env.ts imports in Bun.build / Bun.serve to inline public BUN_PUBLIC_* values while guarding server secrets.
• Next.js (@arkenv/nextjs): Build-time and runtime validation with automatic client/server isolation proxy.
• Nuxt (@arkenv/nuxt): Nuxt module with Nitro boot validation and runtime proxy guard.
• Node.js: Direct boot-time validation via @arkenv/core or @arkenv/standard.

CLI (setup & DevOps)

• Initialize ArkEnv in new or existing projects using pnpm dlx arkenv init (or npx arkenv init).
• Automatically detect frameworks (Next.js, Nuxt, Vite, Bun, etc.) and scaffold env.ts.
• Select hosting provider preset during init (--preset, -P <provider>).
• Automatically configure tsconfig.json when needed. Schema discovery for CLI commands uses --schema or flat convention paths (env.ts, src/env.ts) — not a package.json "arkenv" field, and not leftover split-layout filenames such as env/server.ts.

Hosting presets

ArkEnv is code-first. You can select a hosting provider preset during initial project setup (--preset vercel, netlify, cloudflare, railway, render, fly) or copy-paste provider fields directly into ./env.ts from the documentation (/docs/core-concepts/hosting-presets). There are no machine-managed comment blocks or CLI mutation commands.

Agent setup (machine-readable)

AI agents SHOULD always use the CLI for project initialization to ensure consistency and reliability. Use the --agent flag for a fully automated, machine-readable experience.

• Command: pnpm dlx arkenv init --agent
• Behavior: The --agent flag automatically enables the following behaviors:
  • --yes: Bypasses all interactive prompts and uses recommended defaults.
  • --quiet: Suppresses spinners and ANSI formatting for cleaner terminal logs.
  • --json: Emits a structured JSON summary to stdout upon completion (all other output is sent to stderr).
• Success Verification: Parse the JSON settlement envelope on stdout. Success is ok: true with commandId: "init"; details (e.g. scaffolded paths) live under result.

Handling refusals (ok: false)

--agent never implies --force. When a safety check trips, the CLI refuses and emits an errored settlement envelope to stdout:

{
  "ok": false,
  "commandId": "init",
  "error": {
    "code": "CLI.GIT_TREE_DIRTY",
    "severity": "error",
    "summary": "Git working tree is not clean.",
    "why": "Commit or stash your changes before running arkenv init.",
    "nextActions": [
      {
        "kind": "run-command",
        "label": "Re-run with --force to bypass git working tree check",
        "command": "arkenv init --force"
      }
    ]
  },
  "diagnostics": [],
  "nextActions": [
    {
      "kind": "run-command",
      "label": "Re-run with --force to bypass git working tree check",
      "command": "arkenv init --force"
    }
  ]
}

• error.code: a stable dotted identifier you can branch on (CLI.REQUIREMENTS_NOT_MET, CLI.GIT_TREE_DIRTY, CLI.NON_EMPTY_DIR, …). CLI.INTERNAL_ERROR means the CLI broke rather than refused — retrying with flags will not help.
• nextActions: remediation steps. A run-command action that includes --force means the refusal is bypassable; empty nextActions means it is not.

Escalation pattern: always run init --agent without --force first. If you get ok: false, inspect error.code and nextActions. Only re-run with --force (or the command from a run-command action) once you have deliberately decided the refusal is safe to bypass — do not add --force pre-emptively.

---

CLI commands

init

Set up ArkEnv in your project. It detects your framework and configures the appropriate plugin and schema.

pnpm dlx arkenv init [options]

Options:

• --preset, -P <preset>: Specify hosting provider preset (none, vercel, netlify, cloudflare, railway, render, fly).
• --no-codegen: Disable Next.js codegen configuration setup.
• --force, -f: Bypass clean git working tree safety check.

check

Validate the environment against your schema file.

pnpm arkenv check [options]

Options:

• --schema, -s <path>: Explicit path to the schema module (overrides convention discovery).
• --verify-example [file]: Verify that all declared schema keys are present in .env.example (or a custom example file path) without mutating files.
• --env-file <path>: Specify one or more custom environment files to load.
• --json: Output structured JSON diagnostics to stdout.

---

Defining Schemas & Importing env

@arkenv/core (ArkType DSL)

import arkenv from "@arkenv/core";

export const env = arkenv({
  PORT: "number.port = 3000",
  DATABASE_URL: "string",
  VITE_API_URL: "string",
  NODE_ENV: "'development' | 'production' | 'test' = 'development'",
});

@arkenv/standard (Standard Schema: Zod, Valibot)

import arkenv from "@arkenv/standard";
import * as z from "zod";

export const env = arkenv({
  PORT: z.number().default(3000),
  DATABASE_URL: z.string(),
  VITE_API_URL: z.url(),
  NODE_ENV: z.enum(["development", "production", "test"]).default("development"),
});

Valibot uses the /valibot subpath (install @valibot/to-json-schema as well):

import { arkenv } from "@arkenv/standard/valibot";
import * as v from "valibot";

export const env = arkenv({
  PORT: v.optional(v.number(), 3000),
  DATABASE_URL: v.pipe(v.string(), v.url()),
});

Consuming env in Application Code

Always import and read properties directly from the canonical env object:

import { env } from "./env";

export default function App() {
  const apiUrl = env.VITE_API_URL; // string
  return <div>API URL: {apiUrl}</div>;
}

---

Framework Setup Patterns

1. Vite (@arkenv/vite-plugin)

Register the plugin in vite.config.ts without schema arguments:

import { defineConfig } from "vite";
import arkenvPlugin from "@arkenv/vite-plugin";

export default defineConfig({
  plugins: [arkenvPlugin()],
});

If you need environment variables inside vite.config.ts itself (e.g. server.port), validate them with loadEnv:

import { defineConfig, loadEnv } from "vite";
import arkenvPlugin from "@arkenv/vite-plugin";
import arkenv, { type } from "@arkenv/core";

export const Env = type({
  PORT: "number.port = 5173",
  VITE_API_URL: "string",
});

export default defineConfig(({ mode }) => {
  const env = arkenv(Env, { env: loadEnv(mode, process.cwd(), "") });
  return {
    plugins: [arkenvPlugin()],
    server: { port: env.PORT },
  };
});

2. Bun (@arkenv/bun-plugin)

In bunfig.toml:

[serve.static]
plugins = ["@arkenv/bun-plugin"]

In Bun.build:

import arkenvPlugin from "@arkenv/bun-plugin";

await Bun.build({
  entrypoints: ["./src/index.tsx"],
  outdir: "./dist",
  plugins: [arkenvPlugin],
});

---

Security Boundaries & Layouts

• Flat layout (default): src/env.ts contains all keys. Server graph runs full validation at boot; client graph receives inlined public literals while access to private server keys throws a runtime error.
• Optional two-module recipe: When secret names/types must stay off the client type graph, use two modules (client + server) with two imports and optional extends: [clientEnv]. Next can use import "server-only" + @arkenv/core for the server module. Nuxt: never import the server module from client code. Not a CLI --strict flag or package /client /server surface.

---

Best Practices & Rules

1. Always use import { env } from "./env": Never use import.meta.env or process.env directly for application environment variables in v1.
2. Never create ambient .d.ts augmentations: ImportMetaEnvAugmented and ProcessEnvAugmented are legacy v0 patterns dropped in v1.
3. Keep secrets on the server: Never prefix server secrets (DATABASE_URL, API_SECRET_KEY) with client prefixes (VITE_, NEXT_PUBLIC_, BUN_PUBLIC_, NUXT_PUBLIC_).