sandbox-migrate-to-next
Migrate Cloudflare Sandbox apps from stable @cloudflare/sandbox to @cloudflare/sandbox@next (SDK 1.0 preview). Use sandbox-next for apps already on the preview.
- License 许可
- Apache-2.0
- 最近更新
- 2026-11-14
- Stars
- ★ 1
- Forks
- 0
01 前言 Front Matter
- name
- sandbox-migrate-to-next
- description
- Migrate Cloudflare Sandbox apps from stable @cloudflare/sandbox to @cloudflare/sandbox@next (SDK 1.0 preview). Use sandbox-next for apps already on the preview.
02 资源文件 3
- 许可证LICENSE11.1 KB
- 参考文档references/skill-feedback.md2.8 KB
- 参考文档SOURCE.md418 B
03 SKILL.md 正文
Migrate stable → Sandbox SDK 1.0 preview (@next)
Perform the port. Follow the steps in order. Depth lives in docs—fetch the linked page when a step needs detail.
Human guide: Migrate · 1.0 preview
New projects should start on @next (sandbox-next), not this skill. Day-to-day stable work → sandbox-stable. Deprecated-API cleanup without moving to @next → 2026 deprecation guide first if needed.
Existing apps should migrate when you can, so you are ready when 1.0 becomes the stable release. Do not force production cutover without the user agreeing.
Prefer installed @next types and the migrate doc over memory.
Workflow
- Review hard rules and the replacement map
- Audit the codebase; list hits and target shapes
- Clarify with the user (cutover, bridge, Python image, unclear sites)
- Upgrade package, image, and code
- Validate
Stop after any step that needs a user decision.
Hard rules
- Worker package and container image must be the same
@nextline. - Production cutover uses immediate container rollout. Stable and
@nextcontrol protocols are incompatible both ways; gradual rollout leaves a broken mixed window. In-flight container work can stop. - After cutover,
await sandbox.exec(...)means process started, not command finished. - Argv is as-is (no implicit shell). Shell syntax needs an explicit shell binary.
- Process handles have no stdin → terminals for interactive input.
- Observation
timeout/AbortSignalcancel the wait only, not the process. - No single retry loop for every error.
- Do not invent APIs (
gitCheckouton core, process stdin, string-exec completion helper). - Self-deployed bridge stays on stable (not part of the preview line yet).
Replacement map
| Stable | @next |
|---|---|
SANDBOX_TRANSPORT / transport / setTransport |
Remove — RPC only |
await sandbox.exec("cmd") → buffered result |
await sandbox.exec(argv) → handle, then output / waits |
execStream / startProcess |
Same handle: logs, waitFor*, kill |
| Default / named sessions | Gone — cwd/env per launch, or one shell script |
sandbox.terminal(request) / session terminal |
createTerminal + terminal.connect(request) |
xterm sessionId |
terminalId |
Interpreter methods on Sandbox |
withInterpreter → sandbox.interpreter.* |
gitCheckout |
argv git via exec |
| String kill signals | Numeric only |
Files, mounts, backups, ports, tunnels, proxyToSandbox |
Mostly unchanged (ignore session/transport bits on stable pages) |
Depth: Migrate · after port, day-to-day → sandbox-next
Audit
rg 'SANDBOX_TRANSPORT|transport:|setTransport|enableDefaultSession|createSession|getSession|deleteSession|execStream\(|startProcess\(|killProcess\(|sandbox\.terminal\(|sessionId|gitCheckout\(|SandboxTransport|ExecutionSession'
Also: string exec(, cd then a later exec, bare createCodeContext / runCode on Sandbox.
Clarify (ask when needed)
- OK to cut production with
--containers-rollout=immediate(live processes/terminals/streams may stop)? - Self-deployed bridge? Leave on stable.
- Python interpreter →
-pythonimage variant? - Call sites not covered by the map?
Upgrade
Package and image
npm install @cloudflare/sandbox@next
FROM cloudflare/sandbox:next
# Python: cloudflare/sandbox:next-python
Same prerelease tag on Worker and image when not on floating next.
Code by area
Apply replacements from the map. For each area, implement from the doc—not from stable habits:
| Area | Doc |
|---|---|
| Commands / handles / waits | Processes · Processes API |
cwd / env / secrets |
Environment · Outbound traffic |
| Drop sessions | Migrate · Lifecycle |
| Terminals | Terminals |
| Interpreter | Interpreter |
| Errors | Errors |
| Durable job across requests | Process execution — lifetime / durability |
Commands (shape):
// Before (stable)
const result = await sandbox.exec("npm test");
// After (@next)
const process = await sandbox.exec(["/bin/bash", "-lc", "npm test"]);
const result = await process.output({ encoding: "utf8" });
const server = await sandbox.exec(["/bin/bash", "-lc", "npm run dev"], {
cwd: "/workspace/app",
});
await server.waitForPort(3000, { timeout: 60_000 });
await server.kill(); // numeric; default 15
Terminals (shape):
const terminal = await sandbox.createTerminal({ command: ["bash"], cwd: "/workspace" });
const t = await sandbox.getTerminal(terminal.id);
if (!t) return new Response("terminal gone", { status: 410 });
return t.connect(request, { cursor, cols, rows });
Interpreter (shape):
import { Sandbox as BaseSandbox } from "@cloudflare/sandbox";
import { withInterpreter } from "@cloudflare/sandbox/interpreter";
export class Sandbox extends BaseSandbox<Env> {
interpreter = withInterpreter(this);
}
Git (shape):
const clone = await sandbox.exec(
["git", "clone", "--depth", "1", "--", repoUrl, "/workspace/repo"],
{ cwd: "/workspace" },
);
const result = await clone.output({ encoding: "utf8" });
Delete transport settings entirely. Remove session APIs. Isolate users with separate sandbox IDs.
Deploy cutover
Staging/branch first. Production is one deploy of matching Worker + image:
npx wrangler deploy --containers-rollout=immediate
Leave rollout_active_grace_period at default 0 (or set 0 if raised). After cutover, pre-deploy process/terminal IDs are invalid. Details: Migrate · Container rollouts
Validate
- Lockfile + Dockerfile on the same
@nextline - Typecheck against
@next - Smoke argv
exec+output({ encoding: "utf8" }) - Smoke long process / terminal / interpreter if used
- Errors distinguished: unavailable / interrupted-RPC / stale / local wait
- No live secrets in sandbox env
- Grep again for removed APIs
- Production used
--containers-rollout=immediate
Then day-to-day work uses sandbox-next.
Red flags — stop and fix
- Mixing
@nextWorker with stable image (or reverse) - Gradual container rollout for this cutover
- Treating
await execas command completion - Assuming
cd/ exports persist acrossexeccalls - One retry wrapper for every error
- Inventing
gitCheckout, process stdin, or undocumented APIs - Keeping pre-cutover process/terminal IDs after deploy
- Forcing production cutover without user agreement
- Putting live secrets in
setEnvVars/ launchenv
Skill feedback
If the user reports that this skill gave incorrect or outdated guidance, or is missing information, read references/skill-feedback.md.
04 同仓库更多 Skills
writing-for-agents
Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.
wrangler
Run or troubleshoot Wrangler CLI commands in projects that have a wrangler.jsonc, wrangler.json, or wrangler.toml file, including local development, Previews, deployment, and resource management. For new projects and projects without a Wrangler configuration file, use the cf CLI instead.
workers-profiling
Profile or debug CPU usage and memory allocation in deployed Cloudflare Workers and Durable Objects. Then use this information to optimize your code.
workers-development
Run, inspect, and debug Cloudflare Workers during local development with cf dev, wrangler dev, or the Cloudflare Vite plugin, using Local Explorer to read and seed binding data and query local traces and logs. Also routes production debugging to Workers observability. Use for debugging requests whether or not the project uses Wrangler.
workers-best-practices
Cloudflare Workers best practices for production applications. Use when writing, reviewing, or configuring Workers.
wordpress-pro
Develops custom WordPress themes and plugins, creates and registers Gutenberg blocks and block patterns, configures WooCommerce stores, implements WordPress REST API endpoints, applies security hardening (nonces, sanitization, escaping, capability checks), and optimizes performance through caching and query tuning. Use when building WordPress themes, writing plugins, customizing Gutenberg blocks, extending WooCommerce, working with ACF, using the WordPress REST API, applying hooks and filters, or improving WordPress performance and security.