GitHub Spark in public preview for Copilot Pro+ subscribers

[tuves]

Stuck between idea and implementation? Spending weeks on mock ups or docs that never ship? GitHub Spark takes you from idea to deployed app in minutes.

Build and ship full-stack intelligent apps using natural language with access to the full power of the GitHub platform—no setup, no configuration, and no headaches.

Key features

• Natural language to app: Describe your idea and watch Spark build it, with frontend and backend capabilities included, all powered by Claude Sonnet 4.
• No setup required: Data, LLM inference, hosting, deployments, and GitHub auth all included out-of-the-box.
• Add AI to your apps: Add intelligent features powered by LLMs from OpenAI, Meta, DeepSeek, xAI and more - no API key management needed
• One-click deployments: Publish your app with a single click.
• Build your way: Use natural language, visual editing controls, or code with GitHub Copilot code completions at your fingertips to iterate on your ideas.
• Create a repository: Get a repository with GitHub Actions and Dependabot incorporated in just a click. Everything stays synchronized so you aren't trapped in a sandbox.
• Expand with Copilot agents: Open a codespace directly from Spark to iterate with Copilot agent mode or assign an issue to Copilot coding agent.

Get started

Spark is available in public preview for Copilot Pro+ users, with rollout to additional customers coming soon.

Visit github.com/spark to build your first app.

Copilot Pro+ subscribers receive access as part of their plan. Spark messages use premium requests included in GitHub Copilot plans.

Learn more

• View more details on the Spark product page.
• Read the Spark documentation.

Disclaimer: The UI for features in public preview is subject to change.

[CurationsLA]

Looking forward to testing this out! Just started using GitHub. Great product.

[josekurian]

Great project

[visitoloke3d]

100% Excellent

[Shankar98183235]

GOOD WOOK

[DimitryGarces]

Great project uwu

[askpt]

I was part of the technical preview but I cannot access the github.com/spark only the old link.

[kelseyrae]

Granted you access! Enjoy! We're rolling out to technical preview users as quickly as possible (unfortunately we're limited by model capacity so we have to do a gradual rollout).

[McCrossin]

I have seen the email about this as well but the links don't work for me even though I'm a pro+ user.

[kelseyrae]

Can you share what you're seeing? If you are trying to open the links on mobile and you have the GitHub mobile app on your phone you may need to manually open the link in a mobile browser window.

[McCrossin]

Thanks that's probably the issue just tried checking it out on my commute.

[bij1996]

movies

[mohamedcherifmo]

I just registered to Pro plus and I'm unable to access the repo

[kelseyrae]

Can you share more about what errors you're seeing? You should be able to access it at github.com/spark if you have a Pro+ account.

[guygregory]

Are there any plans to (eventually) make GitHub Spark available to students?

[kelseyrae]

Yes, no concrete date yet, but we are planning to roll out to additional customer segments in the coming weeks and months.

[CurationsLA]

Remember me! Game changer.

[PYRTUS1997]

[pasinduisuranga]

That's great news about GitHub Spark for students! I'm really looking forward to it.

[luyandadhlamini2]

The concept looks great. I've tested it on a sample app. Agent was able to build a usable MVP, but there are teething issues.

Current issues:

1. App keeps refreshing & can't be interacted with.
2. Clicking on publish shows a loading icon next to the word "publishing" & doesn't publish the app.
3. Attempting to open the app in a codespace failed, returning a NETWORK_ERR.
4. Since the app keeps refreshing, feedback can't be provided to the agent to make changes, I assume because the agent waits for app to load before being able to accept input.

My recommendation:
While in preview phase, assign each Spark a UUID (instead of a 20 character name as is currently done) so that we can provide our spark links for the github team to debug & see issues.

@tuves
@kelseyrae
@ghostinhershell

[kelseyrae]

@luyandadhlamini2 Thanks for flagging these! Are you able to share the link to your spark(s) where you were seeing these issues? We can dig in to see what's up.

Regarding the app refreshing - we do load the last stable state of your spark while agentic iterations are in progress. You should be able to interact with your spark during this time, but you won't be able to make other manual changes (in code, data tab, prompts tab, deploy etc). Once the agent finishes your iteration, all of the editing and deploying etc capabilities reactivate.

[mpommrehn]

What is the privacy policy for Spark? Asking relative to a corporate account that includes Copilot (not sure which level), with strict limits on AI usage, only as managed through Copilot under the MSFT Corporate/Enterprise Privacy Policy.
Need to see whether Spark prompts, code, and any generated artifacts are fully contained within the same corporate privacy data wall as the rest of our Copilot usage.
If not now, is that a future intent?
Thanks!

[kelseyrae]

Hi! You can read the data usage policy for all pre-release products like Spark here: https://docs.github.com/en/site-policy/github-terms/github-pre-release-license-terms#4-data-collection-and-usage

If you have any questions please let me know.

[between-sky-sea]

Does spark allow for custom domain?

[kelseyrae]

Not technically, but the URL for your deployed spark is based in part on the name of your spark, which you can change in the settings view.

We automatically manage routing of old domains to new domains if you change the name of your spark after deploying, so you don't need to worry about sticking to a single name of your spark once you deploy.

[mgyarmathy]

Do you plan to incorporate GitHub Organizations into the Visibility settings for deployed apps? Anyone with the link seems quite limited and insecure for internal enterprise apps. @pgahq would be more likely to adopt this if you would enable the use of other Identity Providers (e.g., Okta) instead of Github authentication.

[kelseyrae]

Yes! It's a feature we are actively working on, coming soon.

[mgyarmathy]

Great! We're eager for it!

[agim]

What coding platforms does spark use to build apps?

I asked it to use rails and it decided to use react.

[kelseyrae]

We only support Typescript and React. This is so we can ensure we get users to a deployable app. If you want to use a specific stack, we recommend using Copilot agent mode in VS Code or Copilot coding agent. We see users sometimes start with a prototype of their idea in Spark, then assign an Issue to Copilot coding agent to migrate it to a different language/framework.

[dahatake]

I really want to attach functional requirement document at the first phase. Today we can attach only image to UX/Data prototype.
After initial app is up, I clone to repo to keep work colleague include GitHub Copilot.

[kelseyrae]

Thanks for the feedback! This is also on our radar - coming soon.

[MohammedSardarSaajit4488]

Waiting ::))

[fays3]

Hi bro

[fays3]

Ap mujha Kuch Gide kar sakta hain kya

[MohammedSardarSaajit4488]

Ap mujha Kuch Gide kar sakta hain kya
Yeah what's up??

[mhitza]

Needs a bit more time in the oven. Gave it initial prompt (asked react native, generated some kind of plain react application), didn't implement all the features requested. Asked it to fix a bug, then to implement a separate feature and I got hit with a too many requests error (429). These where 3 requests over the span of what I would call an hour (prompted and went afk for a while each time).

[kelseyrae]

We don't support React Native currently. Only Typescript and React.

Unfortunately it looks like you were trying your request when we were experiencing some overall system capacity constraints (more users were trying to build in Spark than we had available model capacity to support), so some users including you ended up getting rate limited. We've expanded model capacity, so if you'd like to give it another shot you shouldn't run into the same issues.

[himanshu2s]

🔧 Feedback on GitHub Copilot Spark Completion

📌 Subject:

Inaccurate and incomplete completions despite clear and structured prompts

---

🧠 Context:

The backend API is fully operational at [http://localhost:59756/swagger/index.html] and MongoDB is running at [mongodb://localhost:27017]. The frontend (React) was instructed to:

1. Remove all mock data and use the working backend endpoints.
2. Fix the subscription button visibility and interaction bugs.

However, Spark repeatedly responded with messages like:

"All tasks are complete", "Everything is resolved", or "System is working perfectly" — despite nothing being fixed or functional.

---

❌ Core Issues Observed:

• 🔁 False Completion Loops
  Spark asserts success even though the UI remains broken, mock data persists, and real API is not being called.

• 🧩 Lack of Real Integration
  It failed to replace mock logic or wire API calls into the existing UI components.

• 🧪 No Functional Validation
  There was no change in behavior — backend calls were not triggered, nor were frontend bugs fixed.

---

💡 Suggested Improvements:

1. Stateful Prompt Awareness
   Spark should evolve the code based on past prompts and the working system context (like functioning backend endpoints).

2. Backend–Frontend Integration Logic
   When both parts are present, Spark should intelligently connect them, using fetch/axios properly and handling responses.

3. Accurate Task Completion Messaging
   Avoid confirming completion unless issues are demonstrably fixed (e.g., API hits succeed, buttons render properly).

---

✅ Desired Outcome:

A more reliable Spark experience that:

• Replaces mock data with actual API usage,
• Fixes UI rendering bugs (like the disappearing button),
• Connects frontend components to backend endpoints seamlessly,
• Aligns environment configs correctly (CORS, ports, base URLs),
• And avoids misleading "success" messages.

---

Please let me know if code snippets, repo access, or error screenshots are needed for deeper review.

[kelseyrae]

Thank you! Super helpful feedback. If you're able to share any screenshots, link to your repo etc we can take a deeper look.

[KnightNiwrem]

I frequently get the error

Failed to commit AI iteration: Git commit failed: Command failed with code 128: fatal: ambiguous argument 'HEAD': unknown revision or path not in the working tree. Use '--' to separate paths from revisions, like this: 'git <command> [<revision>...] -- [<file>...]'

When creating a new Spark. This prevents further iterations, as the Spark will throw

Failed to submit prompt: Error: Request timeout

when given an iteration prompt.

If I tried to create a repository when it is in this error state, it would simply throw

Failed to fetch

[kelseyrae]

Just an update on this - the team is getting closer to a fix for this. They are actively working on this. Thank you for your patience!

[alexagillman]

@KnightNiwrem Hey! Mind clicking open in codespace and running git commit -m "test" and letting us know what the failure message is?

Any chance you have gpg signing on your commits?

[KnightNiwrem]

@alexagillman The error message is:

@KnightNiwrem ➜ /workspaces/spark-template (main) $ git commit -m "test"
error: gpg failed to sign the data:
[GNUPG:] BEGIN_SIGNING
2025/08/02 01:39:14 error signing commit: error signing commit: error making request: 403 | Current user User does not have push access to this repository, error making request: 403 | Current user User does not have push access to this repository

fatal: failed to write commit object

---

I usually sign my commits and have GPG keys in my account settings.

[alexagillman]

This should be fixed now! Let us know 😄

[KnightNiwrem]

@alexagillman I can confirm that running a couple fresh Sparks (to test) no longer produces the git error.

Will need to wait till they are stale-r to see if they run into the other CORS error.

[jrjfjjtit-netizen]

02097273996

[vs2explore]

Keep getting this error for npm run build command for a private repo in codespace:

failed to load config from /workspaces/note-collector/vite.config.ts
error during build:
Error [ERR_MODULE_NOT_FOUND]: Cannot find package '@github/spark' imported from /workspaces/note-collector/node_modules/.vite-temp/vite.config.ts.timestamp-1754033061716-bf8e325a9910f.mjs

[Jhoel777ar]

Hi, I ran into the same error in my private repo in Codespaces with Vite and @github/spark. I tried a couple of things to fix it, and this is what worked for me, inspired by what I saw in the GitHub Spark thread:

Check if @github/spark is installed correctly:

See if it's in your package.json. If not, add it or verify that it's a valid dependency.
Run npm install to make sure it's in node_modules.
Since it's a private package, I set up a token in .npmrc:
//npm.pkg.github.com/:_authToken=${GITHUB_TOKEN}
https://github.com/github:registry=https://npm.pkg.github.com/
Make sure you have GITHUB_TOKEN as an environment variable in your Codespace.

Check your vite.config.ts:

Verify that the import of @github/spark is correct:
import { defineConfig } from ‘vite’;
import spark from ‘@github/spark’;
export default defineConfig({
plugins: [spark()],
});

If you don't need @github/spark, remove the import and try npm run build again.

Switch to ESM:

The [ERR_MODULE_NOT_FOUND] error gave me headaches until I made sure my project was using ESM.
In package.json, set “type”: “module”.
If necessary, rename vite.config.ts to vite.config.mts.
Change any require to a dynamic import:
const spark = await import(‘@github/spark’);

Clean everything and rebuild:

Delete node_modules and package-lock.json:
rm -rf node_modules package-lock.json

Clear the npm cache:
npm cache clean --force

Reinstall:
npm install

Force the build:
npm run build -- --force

Check your Codespace:

Use Node.js v20 or higher (Vite prefers it).
If you changed your username on GitHub (as I saw in the thread), update the remote:
bashgit remote set-url origin https://<YOUR_PAT>@github.com/<your_username>/<your_repo>.git

Plan B if @github/spark doesn't work:

If @github/spark is exclusive to Pro+ or is not available, replace it with something standard like vite-plugin-react:
import { defineConfig } from ‘vite’;
import react from ‘@vitejs/plugin-react’;
export default defineConfig({
plugins: [react()],
});

Install:
npm install @vitejs/plugin-react

[vs2explore]

wow, great. thanks for detailed instructions. Let me check and get back to you/

[vs2explore]

@Jhoel777ar
I deleted node_modules package-lock.json, getting this error now:

@vs2explore ➜ /workspaces/note-collector (main) $ npm cache clean --force

npm WARN using --force Recommended protections disabled.
@vs2explore ➜ /workspaces/note-collector (main) $ npm install
npm ERR! code E404
npm ERR! 404 Not Found - GET https://registry.npmjs.org/@github%2fspark - Not found
npm ERR! 404
npm ERR! 404 '@github/spark@^0.0.1' is not in this registry.
npm ERR! 404
npm ERR! 404 Note that you can also install from a
npm ERR! 404 tarball, folder, http url, or git url.

[Jhoel777ar]

Ugh, that E404 error with @github/spark is a bit of a problem. It seems that npm is looking for the package in the public registry (registry.npmjs.org), but @github/spark is probably in GitHub Packages or is private.
Configure GitHub Packages:
o I created a .npmrc file in the project root (/workspaces/note-collector):
@github:registry=https://npm.pkg.github.com //npm.pkg.github.com/:_authToken=$ {GITHUB_TOKEN}
o Go to GitHub > Settings > Developer settings > Personal access tokens > Tokens (classic) and generate a PAT with read:packages permissions (and write:packages if you are going to publish).
o Put the token in the Codespace environment variables:
export GITHUB_TOKEN=your_token_here
Or paste it directly into .npmrc, replacing ${GITHUB_TOKEN}.
2. Check package.json:
o Make sure @github/spark is in dependencies:
“dependencies”: { “@github/spark”: “^0.0.1” }
o If you are unsure of the version, check https://npm.pkg.github.com/@github/spark (you must be logged in).
3. Reinstall:
o Clear the cache (as you already did):
npm cache clean --force
o Then:
npm install
4. Check the Codespace:
o Make sure you are using Node.js v20+ (node -v to confirm).
o If you changed your username on GitHub (as I saw in the thread), update the remote:
git remote set-url origin https://<your_PAT>@github.com/<your_username>/<your_repo>.git
5. If @github/spark doesn't pull:
o If the package is only for Pro+ or is unavailable, I removed it from package.json and used vite-plugin-react:
“dependencies”: { “@vitejs/plugin-react”: “^4.0.0” }
I updated vite.config.ts:
import { defineConfig } from ‘vite’; import react from ‘@vitejs/plugin-react’; export default defineConfig({ plugins: [react()], });
I installed:
npm install @vitejs/plugin-react
6. To debug:
o I ran:
npm install --loglevel verbose
to see what happens.
o I also tried:
npm info @github/spark
to confirm if the package exists.
o If it doesn't work, write to GitHub support with your Spark link ([email protected]), as Kelsey Rae said.
If you want to test whether it works or not, run this:
if grep -q “npm.pkg.github.com” .npmrc; then
echo “Everything is fine with the GitHub Packages registry.”
else
echo “Error: Configure .npmrc with @github:registry=https://npm.pkg.github.com”
exit 1
fi
if [ -z “$GITHUB_TOKEN” ]; then
echo “Error: GITHUB_TOKEN missing from environment variables.”
exit 1
else
echo “GITHUB_TOKEN ready. You can run npm install.”
fi

///I hope this helps you.

[Jhoel777ar]

Great initiative with GitHub Spark! To improve the experience, I suggest implementing a more detailed debugging system for errors such as ‘Failed to commit AI iteration’ or ‘Request timeout’. An accessible log with specific steps to resolve them would be useful. In addition, allowing functional documents to be uploaded in the initial phase would streamline collaboration. Keep up the good work!

[chogalesoham]

🔥 What Is GitHub Spark?
GitHub Spark is an AI-powered platform that turns natural language prompts into full-stack applications — instantly. It handles UI, backend, hosting, and deployment without needing any setup or configuration.

Powered by Claude Sonnet 4 (Anthropic), along with other LLMs like OpenAI, Meta, and xAI.

🧠 Key Capabilities
🗣️ Description-to-app via natural language

💾 Built-in data storage, authentication, and GitHub Actions

🚀 One-click deployments and live previews

🔁 Iterate via:

Natural language prompts

Visual tools

Or direct code (with GitHub Copilot integration)

🚀 What’s Included in the Public Preview?
Available to GitHub Copilot Pro+ subscribers
💳 ($39/month or $390/year)

✅ Monthly quota of Spark messages included

🔧 Additional usage may incur extra cost

📱 Build unlimited apps

🧩 Run up to 10 active sessions at the same time

✨ Why Spark Matters
⚡ Speeds up prototyping from weeks to minutes

🛠 No manual config, setup, or API key handling

🌍 Empowers developers and non-developers to bring ideas to life with minimal friction

🛠 How to Get Started
Subscribe to GitHub Copilot Pro+

Go to 👉 github.com/spark

Enter your app idea
e.g., "Build a habit tracker with AI reminders"

Let Spark generate the full-stack app

Refine via:

💬 Natural language

🧱 Visual editor

👨‍💻 Codespaces

Click Deploy — Spark handles hosting, GitHub integration, and CI/CD!

[mochan-tk]

If this is information you can share, could you please let me know how to obtain the SNAPSHOT_SAS_URL?
https://github.com/github/spark-template/blob/main/.devcontainer/postStartCommand.sh#L21

[mochan-tk]

By the way, do Codespaces launched via GitHub Spark not appear in the management dashboard?
https://github.com/codespaces

[nkasifshaike-byte]

Stuck between idea and implementation? Spending weeks on mock ups or docs that never ship? [GitHub Spark](https://github.com/spark) takes you from idea to deployed app in minutes.

Build and ship full-stack intelligent apps using natural language with access to the full power of the GitHub platform—no setup, no configuration, and no headaches.

Key features

• Natural language to app: Describe your idea and watch Spark build it, with frontend and backend capabilities included, all powered by Claude Sonnet 4.
• No setup required: Data, LLM inference, hosting, deployments, and GitHub auth all included out-of-the-box.
• Add AI to your apps: Add intelligent features powered by LLMs from OpenAI, Meta, DeepSeek, xAI and more - no API key management needed
• One-click deployments: Publish your app with a single click.
• Build your way: Use natural language, visual editing controls, or code with GitHub Copilot code completions at your fingertips to iterate on your ideas.
• Create a repository: Get a repository with GitHub Actions and Dependabot incorporated in just a click. Everything stays synchronized so you aren't trapped in a sandbox.
• Expand with Copilot agents: Open a codespace directly from Spark to iterate with Copilot agent mode or assign an issue to Copilot coding agent.

Get started

Spark is available in public preview for Copilot Pro+ users, with rollout to additional customers coming soon.

Visit github.com/spark to build your first app.

Copilot Pro+ subscribers receive access as part of their plan. Spark messages use premium requests included in GitHub Copilot plans.

Learn more

• View more details on the Spark product page.
• Read the Spark documentation.

Disclaimer: The UI for features in public preview is subject to change.

go bakar

[Joseph85892]

Warking online

[itamarkra]

I'm struggling with manually changing my app. I want to revert changes Spark has made, but every time I force push my git reset, Spark replaces the main branch back to where it was before.
That means, the "two way binding" doesn't really work. It always takes "Spark's" version rather than taking the repository as the single source of truth.
How can this be fixed?

[pasinduisuranga]

What are the benefits for me?

[itamarkra]

How does one deploy the app for production? Or is it just a prototyping mechanism?
I cannot see a way to deploy it with my own domain.

[MasayukiOzawa]

When you published, a rate limit of 429 occurred.
In this discussion, we are discussing the content of this article, but is there anything that users can do to address this issue?

[itamarkra]

Another issue - when Spark responds to a request, it outputs its logic. After it finishes, I cannot seem to be able to restore Spark's logic.
Is there a way to do it? Or can you please add this feature?

[itamarkra]

Another related feature is something that's very powerful with Base44 - discuss mode.
It allows to ask a question without the bot actually programming anything. A brainstorming or design session before implementing.
Can this mode also be added?

[hiarun01]

amazing project

[askpt]

I have been testing it in Edge but it seems it is not working in Safari. Tried to look into the release notes and cannot find any report stating it is not supported in Safari on macOS. Is that expected?

[MasayukiOzawa]

It appears that the GitHub Spark limit has been reached, and the following message is now displayed.

I couldn't find any information about rate limits. How much usage will result in restrictions?