Migrating To Workflow Sdk logo

Migrating To Workflow Sdk

OrganizationPopular
vercel
migrating-to-workflow-sdk

Migrates Temporal, Inngest, Trigger.dev, and AWS Step Functions workflows to the Workflow SDK. Use when porting Activities, Workers, Signals, step.run(), step.waitForEvent(), Trigger.dev tasks / wait.forToken / triggerAndWait, ASL JSON state machines, Task/Choice/Wait/Parallel states, task tokens, or child workflows.

Overview

Publishervercel
Repositoryworkflow
Skill namemigrating-to-workflow-sdk
Stars
2.4K
Forks
373
Bundled files
13
LicenseApache-2.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 13 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by vercel on GitHub. Read the source before you install it.

Installation

Install the Migrating To Workflow Sdk AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/vercel/workflow/tree/main/skills/migrating-to-workflow-sdk
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/vercel/workflow.git /tmp/workflow
mkdir -p .claude/skills
cp -r /tmp/workflow/skills/migrating-to-workflow-sdk .claude/skills/migrating-to-workflow-sdk
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Migrating To Workflow Sdk in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Migrating To Workflow Sdk on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Migrating To Workflow Sdk is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Migrating to the Workflow SDK

Use this skill when converting an existing orchestration system to the Workflow SDK.

Intake

  1. Identify the source system:
    • Temporal
    • Inngest
    • Trigger.dev
    • AWS Step Functions
  2. Identify the target runtime:
    • Managed hosting -> keep examples focused on start(), getRun(), hooks/webhooks, and route handlers.
    • Self-hosted -> also read references/runtime-targets.md and explicitly say the workflow/step code can stay the same, but deployment still needs a World implementation and startup bootstrap.
  3. Extract the source constructs:
    • entrypoint
    • waits / timers
    • external callbacks / approvals
    • retries / failure handling
    • child workflows / fan-out
    • progress streaming
    • external side effects

Default migration rules

  • Put orchestration in "use workflow" functions.
  • Put side effects, SDK calls, DB calls, HTTP calls, and stream I/O in "use step" functions.
  • Use sleep() only in workflow context.
  • For Signals, step.waitForEvent(), and .waitForTaskToken, choose exactly one resume surface:
    • resume/internal -> createHook() + resumeHook() when the app resumes from server-side code with a deterministic business token.
    • resume/url/default -> createWebhook() when the external system needs a generated callback URL and the default 202 Accepted response is fine.
    • resume/url/manual -> createWebhook({ respondWith: 'manual' }) only when the prompt explicitly requires a custom response body, status, or headers.
    • If a callback-URL prompt does not specify response semantics, default to resume/url/default and make the assumption explicit in ## Open Questions.
  • Never pair createWebhook() with resumeHook(), and never pass token: to createWebhook().
  • Wrap start() and getRun() inside "use step" functions for child runs.
  • Use getStepMetadata().stepId as the idempotency key for external writes.
  • Use getWritable() in workflow context to obtain the stream, but interact with it (write, close) only inside "use step" functions.
  • Prefer rollback stacks for multi-step compensation.
  • Choose app-boundary syntax in this order:
    1. If the prompt explicitly asks for framework-agnostic app-boundary code, use plain Request / Response even when a framework like Hono is named.
    2. Otherwise, if the target framework is named, shape app-boundary examples to that framework.
    3. Otherwise, keep examples framework-agnostic with Request / Response. Do not default to Next.js-only route signatures unless Next.js is explicitly named.

Fast memory aid:

  • Callback URL + default ack -> createWebhook()
  • Callback URL + custom ack -> createWebhook({ respondWith: 'manual' })
  • Deterministic server-side resume -> createHook() + resumeHook()

Fast-path router

Load references/resume-routing.md when the source pauses for Signals, step.waitForEvent(), or .waitForTaskToken.

Fast defaults:

  • callback URL only -> resume/url/default
  • callback URL + explicit custom response -> resume/url/manual
  • deterministic server-side resume -> resume/internal
  • self-hosted -> add runtime/self-hosted
  • named framework -> add boundary/named-framework
  • explicit framework-agnostic request -> add boundary/framework-agnostic

Before drafting ## Migrated Code, write the selected route keys in ## Migration Plan.

Source references

  • Temporal -> references/temporal.md
  • Inngest -> references/inngest.md
  • Trigger.dev -> references/trigger-dev.md
  • AWS Step Functions -> references/aws-step-functions.md

Shared references

  • references/shared-patterns.md: Reusable code templates for hooks, child workflows, idempotency, streaming, and rollback.
  • references/runtime-targets.md: Managed vs. custom World guidance.
  • references/resume-routing.md: Route-key selection, obligations, and exact ## Migration Plan shape.
  • references/retries.md: Canonical retry mechanics, including stepFn.maxRetries, RetryableError({ retryAfter }), and FatalError.

Required output shape

Return the migration in this structure:

md
## Migration Plan
## Source -> Target Mapping
## Migrated Code
## App Boundary / Resume Endpoints
## Verification Checklist
## Open Questions

Verification checklist

Fail the draft if any of these are true:

  • ## Migration Plan omits Route keys
  • ## Migration Plan omits Why these route keys
  • ## Migration Plan lists route keys that do not match the prompt
  • ## Migration Plan lists required code obligations that do not match the selected route keys
  • Source-framework primitives remain in the migrated code
  • Side effects remain in workflow context
  • sleep() appears inside a step
  • Stream interaction (getWriter(), write(), close()) appears inside a workflow function
  • Child workflows call start() / getRun() directly from workflow context
  • External writes omit idempotency keys
  • Hooks/webhooks are missing where the source used signals, waitForEvent, or task tokens
  • A callback-URL flow uses createHook() + resumeHook() instead of createWebhook()
  • A resume/url/default or resume/url/manual migration invents a user-authored callback route or resumeWebhook() wrapper when webhook.url should be the only resume surface
  • createWebhook() is given a custom token or paired with resumeHook()

Validation note:

  • Reading webhook request data in workflow context is allowed. Only request.respondWith() is step-only.

Additional fail conditions:

  • resume/internal output omits resumeHook() in app-boundary code
  • resume/internal output omits a deterministic business token
  • resume/internal output emits createWebhook() or webhook.url
  • resume/url/default output does not pass webhook.url to the external system
  • resume/url/default output emits resumeHook(), respondWith: 'manual', or RequestWithResponse without a custom-response requirement in the prompt
  • resume/url/default output invents a user-authored callback route or resumeWebhook() wrapper when webhook.url is the intended resume surface
  • resume/url/manual output does not pass webhook.url to the external system
  • resume/url/manual output omits RequestWithResponse or await request.respondWith(...)
  • resume/url/manual output calls request.respondWith(...) outside a "use step" function
  • resume/url/manual output invents a user-authored callback route or resumeWebhook() wrapper when webhook.url is the intended resume surface
  • createWebhook() is paired with resumeHook()
  • self-hosted output omits World extends Queue, Streamer, Storage, startWorkflowWorld(), or the explicit note that the workflow and step code can stay the same while the app still needs a custom World
  • named-framework output mixes framework syntax with plain Request / Response app-boundary code without a framework-agnostic override

For concrete passing code, load:

  • references/shared-patterns.md -> ## Generated callback URL (default response)
  • references/shared-patterns.md -> ## Generated callback URL (manual response)
  • references/runtime-targets.md -> ## Self-hosted output block
  • references/aws-step-functions.md -> ## Combined recipe: callback URL on self-hosted Hono

Sample prompt

text
Migrate this Inngest workflow to the Workflow SDK.
It uses step.waitForEvent() with a timeout and step.realtime.publish().

Expected response shape:

md
## Migration Plan
## Source -> Target Mapping
## Migrated Code
## App Boundary / Resume Endpoints
## Verification Checklist
## Open Questions

Example references

Load a worked example only when the prompt needs concrete code:

  • references/shared-patterns.md -> ## Named-framework internal resume example (Hono)
  • references/shared-patterns.md -> ## Generated callback URL (default response)
  • references/shared-patterns.md -> ## Generated callback URL (manual response)
  • references/runtime-targets.md -> ## Self-hosted output block
  • references/aws-step-functions.md -> ## Combined recipe: callback URL on self-hosted Hono

Reject these counterexamples:

  • resume/url/default or resume/url/manual + user-authored callback route when webhook.url is the intended resume surface
  • createWebhook() paired with resumeHook()
  • named-framework app-boundary output mixed with plain Request / Response without a framework-agnostic override

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Migrating To Workflow Sdk AI skill do?

Migrates Temporal, Inngest, Trigger.dev, and AWS Step Functions workflows to the Workflow SDK. Use when porting Activities, Workers, Signals, step.run(), step.waitForEvent(), Trigger.dev tasks / wait.forToken / triggerAndWait, ASL JSON state machines, Task/Choice/Wait/Parallel states, task tokens, or child workflows.

Why use Migrating To Workflow Sdk on TypingMind?

Because you install it once and use it with any model. Migrating To Workflow Sdk is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Migrating To Workflow Sdk in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/vercel/workflow/tree/main/skills/migrating-to-workflow-sdk. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Migrating To Workflow Sdk?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Migrating To Workflow Sdk?

As many as you like. As long as a model supports skills, you can use Migrating To Workflow Sdk with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Migrating To Workflow Sdk AI skill free?

Yes. It is published on GitHub by vercel under the Apache-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇