#opal-show-n-tell・🖼️

1 messages · Page 1 of 1 (latest)

ruby sinew
#

first

frosty dust
#

👀

gusty sequoia
#

Third

woven pier
#

Four™

worldly cobalt
#

∞🤚

silk magnet
#

Go 💖 Opal, GO!

earnest thicket
#

Hi Opalites! Please share Opals you're making to help inspire others!

marble panther
#

how to enable mcp in opal?

velvet flicker
#

[Bug / Error?]

  1. Sometimes the [Generate] block shows a page saying "no items".

  2. After repeating the [Generate] process many times, a static page (with no input field) appears showing only the named result of the [Generate] block, and I can’t proceed any further.

  3. When I click [Share App] → [View Share Permissions], the screen often freezes.

This link is the Opal app I tested.
https://opal.withgoogle.com/?flow=drive:/1bC-R7y6wLzkodssBzqgady9qgJ1tyo9x&shared&mode=app

ashen orbit
tulip pawn
ashen orbit
#

Hey, do you guys happen to have a killer one-shot prompt for Opal? 👀 would love to see what you’ve got haha

obtuse lava
#

hi from vietnam

ruby turtle
#

hi, do you have a way to control output to files into a G.Drive folder?

frozen grotto
#

hii

weary night
royal lintel
#

I still use one of the first ones I made. "Become Super". You can upload a photo of yourself and enter a short superpower wish, it generates a superpower card with a new photo and Veo video. One of my favorite things I've made 😂 that says a lot

ashen orbit
#

Mobile experience is not good yet

paper sky
royal lintel
#

I just used a Midjourney image for the starting image on this one, but you can try any. Not the greatest superpower with this one 😂 I still need to go back and fix the save image/video thing.

royal lintel
earnest thicket
royal lintel
earnest thicket
random wyvern
random wyvern
#

Another Opal creation with my app 3D print master. It generates stunning 3d animations based on your image and style input.

random wyvern
#

I was in my room looking in mirror, and a Nun appeared. Made with my Opal App called 3D Print Master.

paper sky
paper sky
royal lintel
random wyvern
#

I just made another app with opal for YouTube channels. Not only does the app make a perfect intro, but is makes a whole brand kit for your channel whatever it might be about. It provides you with the following items, short script, call to action lines, branding color presets, WCAG Contrast Accessibility Report, thumbnail image to use, and audio to go with intro. just follow instructions in app and that's it . I included a sample video I made for my channel Monkey Moments that has videos and stories about baby monkeys life. This my share link so you can try app for yourself. I'll share as soon as i figure out why share link stopped working.

golden wing
#

@random wyvern Something like this is perfectly fine to share

random wyvern
weary night
ruby sinew
#

fwd from @random wyvern 🙂

weary night
#

this is the closest I've gotten to perfect
I want cards to still be able to expand and show all details even when on cooldown, or even when defeated, and to be able to be revived if accidentally marked defeated or if revived by a card ability. and now I will re-roll the app but some other part of the app might break like the deck builder screen might be empty or none of the + buttons work or hover windows won't close lol
https://opal.withgoogle.com/?flow=drive:/10gMsFspThgfldZXQJ-KdNxCNrDPoG-V1&shared&results=103Lbze1sopeFZljTGrt1bt6DSaPNpvEJ&mode=app

jagged moon
weary night
#

2 players play on separate phones in person, using thier devices to track their own cards and HP

random wyvern
random wyvern
fallow owl
random wyvern
grand mirage
#

I get sharing these with eachother and stuf but what about people otherr than developers? People that don't carea bout the guts.

frigid bridge
#

Curious if there's a place where people share their creations. You know how there's the Gallery. Are there others?

frigid bridge
grand mirage
worldly wharf
#

Friends i need your suggestions, Based on the user input, i wanted to have two paths If (This) then ... else .... How to do this in Opal? Also is there any way to add buttons in the UI page? Thanks for your help

grand mirage
#

could the code execution tool do that for you? Or am i off base haha

worldly wharf
#

How to add the code execution tool in Opal to clasify the input

worldly wharf
grand mirage
# worldly wharf Can u pls suggest how to clasify the input and decide the path of execution in o...

I havent played with that option much but i got it to kind of work if i tell the AI that if it's designated path isnt included in the last step than it doesnt need to proceed. I created a route message node with a code execution with an example code im not sure if it even worked you cant really see what is being passed between the nodes like in n8n. I'll share what i have so far with you and maybe you can build off of it

worldly wharf
grand mirage
#

Well since it shows my prompts and stuff for certain ones I haven't really made it public but once they add the option to just let people use it then I will share the whole arsenal lol

paper sky
#

I’m tinkering with a landing page in the wee hours of the night…any recommendations on integrating a google form?

weary night
#

The app seems to be about to ask users for input, and I think they said it can write to a spreadsheet but it can read the spreadsheet, so really it can't do more right now than a form can already do, unless you want to add logic or calculations to the users answers before adding them to a spreadsheet. (That is if I understand the functionality as it stands now, I've only read from it, but it only reads the spreadsheet once when building the app, then it can't access it anymore after that.)

worldly wharf
#

When we share the app to other users, it is asking permission for their google drive, some are not ok to give all the permission to view the app, any alternative way to share the app where it does not ask all the permission to view or try the app

dense holly
#

I'm in South Korea. I'm building a UGC content creator app using Nanobanana, Script, Lyric and Veo3 via Opal.

But quota limited error has not been solved. Is it occurred because of the region? or just daily limit?

acoustic cove
#

Hi all, I recreated a job match app I had previously built using chatgpt with Opal last week. I was fascinated by the tool. While there are some improvements needed, I was impressed for two reasons: 1. Opal has lots of potential 2. It actually worked and found two jobs I had not found before.

A quick write up on my experience here: https://craftedforscale.substack.com/p/rebuilding-the-job-bot-in-opal
To play with the job bot go to the link, or here: https://opal.google/?flow=drive:/1uPzqPE9ihSfC_VSJ_OvJ1KO7nFKo9ZmC&shared&mode=app

I rebuild the job bot in Opal to surprisingly good results

whole creek
acoustic cove
acoustic cove
earnest thicket
#

we're updating the sharing flow soon to make it more clear, but once you publish an app, you have to choose to push updates! So if you click "share app" you might see this: "Your app has unpublished changes" next to the an option to click "publish". Hope that fixes it and makes sure your users get the latest version @acoustic cove

#

I'm in South Korea. I'm building a UGC

acoustic cove
earnest thicket
#

ah got it, worth checking! Sorry you hit that bug! If you see it again, let us know

acoustic cove
#

No worries, excited to see this tool developed. Lots of potential!

paper sky
earnest thicket
#

It worked for me, just tested it! FYI @acoustic cove , very cool results and love the ranking! Try it again @paper sky ? We had a model outtage for Flash 2.5 yesterday I believe, so that may have been the issue (unrelated to the Opal).

acoustic cove
acoustic cove
chrome prism
#

I'm still trying to figure why some of the apps in this thread open just fine. While others you're forced to sign into your Google account.

iron void
iron void
earnest thicket
trail sparrow
thin musk
#

Hi

fringe briar
#

hello

bronze sparrow
#

Not working...

crude bough
#

📢 Hello Orca/Opal Studio Engineering Team,

We are reporting a critical technical issue related to our workflow deployment in the Orca/Opal Studio environment.

🧩 1. Issue Description
Category Detail
Model Used Gemini 2.5 Pro
Platform Orca / Opal Studio
Core Issue The model is forcefully outputting its internal reasoning process (the thought log, e.g., thought Here's a thinking process...) immediately before the final user response.
Impact This behavior severely degrades user experience and breaks the persona of our agent “Yeonseo,” who is designed to be warm and human-like.
⚙️ 2. Attempts at Resolution (Front-end Control)

We have attempted to block the log output using every available command within the Advanced Settings / Prompt field, but these attempts have been unsuccessful.

A. Critical SYSTEM RULE Commands

SYSTEM RULE

!! CRITICAL: NEVER output any reasoning, thought, analysis, or debug trace.

B. Explicit Text Blocking
!! Do not use <thought>, <reasoning>, or "Here's a thinking process" text.

C. Generation Config Cleanup

Removed the “Reasoning: ✅ Allowed” statement from Technical Settings.

crude bough
# crude bough 📢 Hello Orca/Opal Studio Engineering Team, We are reporting a critical technic...

🧠 3. Request for Solution (Backend Parameter Adjustment)

Since prompt-level control has proven ineffective,
we believe this requires a backend adjustment to the model’s generation parameters applied to our “Answer Fan Question” node (which calls Gemini 2.5 Pro).

We kindly request that the following parameters be explicitly set:

Parameter Recommended Value Purpose
thinkingBudget 0 Prevent the generation of any internal reasoning stage. (We understand that in Gemini 2.5 Pro, setting this to 0 may require backend override.)
includeThoughts false Ensure that even if thoughts are internally generated, they are not included in the final user response (secondary safeguard).
💬 4. Expected Outcome

Applying the above parameters should completely suppress any visible reasoning output, restoring a clean, natural, persona-based dialogue experience for users.

🙏 5. Appreciation

We greatly appreciate your help in maintaining a consistent, human-like persona for our users.
Thank you so much for reviewing this — it’s critical to our live deployment stability.

(Optional Addendum for Context)

Our agent “Yeonseo” is a narrative-driven character.
Any unintended reasoning or “thought” log breaks immersion, appearing as raw debug text to the user.
We’ve already verified that the same prompt performs normally under GPT-5 (no reasoning leaks), so this issue appears specific to Gemini-based or Opal-rendered pipelines.

night siren
#

How to publish our opal without showing the editor tab?

earnest thicket
night siren
chilly perch
#

How to update google sheet fields and columns with opal?
Hello, I am trying to create a social media automation, thought I am able to create the post fetching input from google sheet, I am unable to update it in the same sheet.

Basically, the app shows the output but fails to update, is there need for any sheet api to be enabled to make it work?

can anyone help me wit this?

river storm
#

**How to update google sheet fields and

karmic atlas
#

thanks for the demo @earnest thicket yesterday!

  • used the recipe idea to test outputs as inputs, and cool to see it work
  • for some reason i thought it was closed loop (ie. only one input and one output for opal)
  • this interaction is so fun! bringing human in the loop at different parts of the process

here's video demo

https://www.loom.com/share/494c87253092421996692f111f38c625

here's proof of concept for anyone in case it helps
https://opal.google/?flow=drive:/1GmJSnJ4_xDaYY8ce23syvnQePyUVuZ2b&shared&mode=app

keen stratus
#

For a Google Labs Make-A-Thon, I made a resume optimizer!
Thought I would share it here since i myself take a long time to try to optimize my resume for a specific job. It gets you through the ATS sensors via extracting key words in the job description and altering your resume to be tailored.
It does a check at the end to ensure that everything fits in one page. Edits your resume doc directly

https://opal.google/?flow=drive:/10_Ukc52AnYk_PskdV20GTADSh1pmHZoZ&shared&mode=app

earnest thicket
keen stratus
faint anvil
#

I figured out how to have Opal dynamically update values in a Google Sheet with other data and formulas even though Opal can only append data at the end of a Google Sheet as far as I know.

I created several apps that turn screenshots from a mobile game into tables that I then am able to run formulas to figure out additional things like do I have enough resources to build the next building.

The key is to have the output go to an intermediate file where Opal can append the data and then use the

=IMPORTRANGE("url of google sheet", "sheetname!A"&IMPORTRANGE("url of google sheet", "sheetname!E1")&":B"&IMPORTRANGE("url of google sheet", "sheetname!E1")+6)

This returns the 2 columns of last 7 rows from the spreadsheet that Opal outputs to
E1 contains a =COUNTA(A:A)-6 on that sheet to determine what the starting row to import from is. [Note: Numerous AI chatbots wanted to do a sort on the date to get the last 7 rows assuming it was sequential.]

You have to approve the connection of the Google Sheets for it to work and you need to have blank cells where the data is going to go because the IMPORTRANGE function will not overwrite existing data.

river storm
#

I figured out how to have Opal

faint anvil
#

Following the tutorial idea, I made a menu planner customized for my family that takes likes, things to avoid and recent menus from docs (so it is easy for anyone in family to adjust) and generates a custom set of menus with the detailed recipes followed by a combined shopping list for the week. The menu is appended to a past meals list which feedbacks into the mini app to avoid repeats or very similar meals with recent ones.

I’m on my 1st week using this as an alternative to Hello Fresh and the first meal turned out great. Next step is to add some sort of rating system to up vote or down vote meals we make.

https://docs.google.com/document/d/1qFwq3WYYF5-Aa7SQHGUp7qpnbOebRUtwdrEghXJt4c0/edit?usp=drivesdk

brazen salmon
velvet pollen
keen stratus
# subtle crow oops

Hi, thats likely occurring because the site you put in doesnt allow robots to scrape which Gemini honors. If that occurs you can paste in the job description!

frosty lily
#

Hello @errant crown we wanted to remind you of our server guidelines, as stated in #start-here・📍 : “Currently, our team can support engagement in English; the discussion in this server will largely remain in English for now.” Thank you.

cobalt helm
#

Hi! A friendly reminder videos and images cannot be created within the server. Please visit https://labs.google/ and pick a tool to get started. If you need help please make sure to ask clearly and do so in the correct channel. Thank you!

quasi goblet
#

Here is a simple Opal template i wanted to share that takes Agile Software Development principles and applies them to Agentic AI. Basically give it a problem and a number of agents to do said problem and let it build a phased plan to assign work to each agent. Try it out, remix it for your architecture, and add a "prompt engineer" step at the end to prepare the work for your agent.

https://opal.google/_app/?flow=drive:/1_QaYrnASoCeXi4_vfEX_yKjBrKAvCWkp&shared&mode=app

river storm
empty ivy
#

Is there an issue publishing apps?

potent whale
river storm
#

Is there an issue publishing apps?

broken roost
#

i just cant use it any more bc of reaching limits google help me

past atlas
#

Subject: [Blocker] Persistent 500 Error (Auto-layout) & Input Echoing (Manual Mode) with Large JSON Input

Hi team, I’m encountering a critical issue with an LLM node (Final Report Viewer) that processes a large JSON object generated from a previous step. I've been stuck on this for a while and need insights.

Context:
-Input: Large JSON object containing a markdown_report key.
-Goal: Extract the markdown text and render it as styled HTML.

The Issues:

  1. Webpage with Auto-layout Mode → 500 Internal Error
    -Whenever I enable this mode, the node crashes with a 500 Internal Error.
    -Attempts: I tried stripping ALL HTML tags, styles, and even reducing it to plain text. The 500 error persists. It seems like the Auto-layout engine crashes immediately upon receiving this specific large data payload.

  2. Manual Layout Mode → Model "Echoes" Input
    -To bypass the 500 error, I switched to Manual Layout.
    -Symptom: The LLM completely ignores the prompt instructions (System Override, Negative Constraints, Extraction Rules) and simply returns the raw input JSON string as the output.
    -Attempts: Tried "Extraction Mode" prompts, specific target keys, and explicit "DO NOT output JSON" rules. Nothing works; it just echoes the input.

Hypothesis:
It seems like the input token limit or context window is being exceeded, causing the platform to crash (Auto-layout) or the model to default to echoing (Manual).

Question:
Is there a known hard limit for JSON input size for these nodes? Or is there a specific workaround to force the model to process the data instead of echoing it when the payload is large?
Any help would be appreciated!

earnest thicket
#

Subject: [Blocker] Persistent 500 Error

craggy basin
#

I build a blog writer, Does the current OPAL not have custom API node? I was planning to automatically publish through hashnode API

cold topaz
gloomy yarrow
worn copper
fading current
#

Hi fam,
Is there any way you can publish opal but without giving an option to remix or see the prompts used ? Or api which will use trigger opal flow through another code base ?

brazen salmon
cold topaz
#

Yeah sometimes the imdb links work and sometimes they don't

#

I am not sure how to get like a accuracy loop to confirm the link is the correct source

vestal chasm
patent hill
#

I can’t access even what show my backend handmade files

patent hill
#

Ans please

#

Hi

fluid trail
#

<@&1366831844502274180> these links ask permission to download all of my Google drive

@valid anvil what’s best way to tag mods?

native rapids
rocky yarrow
fast notch
warped elbow
noble moat
#

I would like to be able to print the web output generated by Opal app. Has anyone be able connect and print?

dusk ore
#

I dont know what im doing

#

Was trying to make a navigation app for apps and well online information refrence like cutting through all the text noise to get straight answers to normally buried context i guess but idk if its ideal or practically useless in regards to current ai models 🤔 most people would just ask ai i assume

#

Maybe better used for ai if optimized as external resource? Like exterior pre processing??? Idk probably a stretch...

quick pond
jolly bear
#

Anyone know if you can use opal as a workflow for stitch?

river storm
#

Anyone know if you can use opal as a

bold bloom
#

Can we only use Gemini 2.5 Flash in Google Opal. Are there limits on other models?

jolly coyote
#

What do I need to do to use it in Türkiye? I came to Türkiye for work and I want to use and experience this AI, but the system recognizes me as living in Türkiye and doesn't accept me. What should I do?

vernal sequoia
#

Hello! I've created an app in Opal to help my wife and I document and better understand what's going on with our new dog who is super fearful and anxious: https://opal.google/?flow=drive:/1n8mTVSN5yx6KSvzKRpl2zkbmEMyFp_e9&shared&mode=app

Something I'm a bit confused about is the initial name field is set to a text element, but still offers an upload option - can this be change to text only? It would be even better if there was an option for a dropdown menu or checkbox instead, but I understand this isn't an option currently.

We take the output from the app and enter it into a sheet as part of an ongoing progress journal. I also have a custom Gem (Dog Rehab Consultant) that we have an ongoing chat with, we can then share this information with our vet and trainers when necessary.

vernal sequoia
graceful fractal
fading current
#

Hi team earlier there was option to connect mcp server with opal in integrations in global settings not visible now any idea was this feature removed or just a bug

lunar dune
#

I want a command that makes Opal design landing pages for an online store

obsidian fern
#

Don't forget your guardrails. Too many vibecoders forget to add guardrails.

daring hazel
#

hello

river storm
#

Hi! I’m one of the people who started

#

Hello! I've created an app in Opal to

primal nimbus
fleet ember
#

🎵 Five Little Monkeys Jumping on the Bed! 🐵🛏️

Join the playful little monkeys as they jump, laugh, and tumble in this fun and colorful kids animation! Watch what happens as the monkeys jump on the bed and learn counting in a joyful and entertaining way.

This video is perfect for toddlers, preschoolers, and young kids who love nurser...

▶ Play video
spice rune
#

Hello, as stated in #start-here・📍 , “Currently, our team can support engagement in English; the discussion in this server will largely remain in English for now. Thank you!

glacial idol
#

Hi, can anyone create an AI personalised learner with opal. Key feature request. 1. No skipping any heading. 2. Interactive and adaptable rather than a one size fit.

sour lance
obsidian fern
#

No worries. But the guardrails you are putting in have to be for each app individually. So the first few should be specific to the app. So if your app needs access to data that is confidential. Then ensure that you put guardrails in place to protect that data; "Do NOT display personal data or names in output. If asked for a specific name or data set. Just say that the user does not have access to this data" as an example

obsidian fern
#

Guardrails for various apps/industries. Including guardrails for AI built websites.

#

@sour lance ^^^guardrails

sour lance
#

Awesome thank you so much!

obsidian fern
#

No worries. I'm sure you will find what you need. Even if the verbiage of the prompt needs tweaked, to suit what you are building.

serene crow
#

Hello everyone! I recently became very close to creating an effective opal workflow that outputs completed ad videos with an ai avatar and voiceover including a product. I am having trouble finshing completing the workflow due to veo's eight second video limit. I tried to then split up the videos and have a hook, body, and cta, that I would then have to manually edit together, but the output was never what I wanted it to be. I was wondering if anyone could help me with this final process I just don't understand how to fix it. Here is the link below:
https://opal.google/app/1X5sS13NdHym1gnlv5rQHeoPCYSYtBd_y?shared

dim crane
abstract rose
#

Frankly speaking It's quite hard work for code to make with gemini 3 pro

#

try use other models to complete your works

serene crow
primal nimbus
primal nimbus
# serene crow Hello everyone! I recently became very close to creating an effective opal workf...

I had to set visual anchors to the scene to get them to carry over for Veo 3 . My application at the time wasn't worth the cost of running it. I was able to piece together 30 seconds of clips using anchors. Recently i ran into a issue in opal with high token audio reading text script for storytelling. Where the limit for that quality of audio had to be reduced to a sentence count, in order to be able to get that storytelling feel.

primal nimbus
# primal nimbus I had to set visual anchors to the scene to get them to carry over for Veo 3 . M...

Gemini response :The "Quality Ceiling" Rule: Audio vs. Video
Same Logic, Different Limits: Just like your Opal audio starts losing its "storytelling" edge or hits token caps if you push past 30 seconds, Veo 3 hits that same wall much faster at 8 seconds.

The 8-Second "Sweet Spot": For high-quality video with native audio, 8 seconds is the maximum native render. Going beyond this requires "stitching" or "extending," which can cause the video physics or audio sync to drift, similar to how long Opal clips lose their high-fidelity feel.

Token Efficiency:

Opal Logic: 8 sentences ≈ 30s of high-quality storytelling audio.

Veo 3 Logic: 8 sentences ≈ 8s of high-fidelity video + synced audio.

Why the limit? To maintain "Cinematic Realism" (4K upscaling and 24fps), the model focuses all its "attention" on that short window. Keeping your prompt condensed to those 8 sentences ensures the AI doesn't run out of "logic tokens" for the physics and sound.

charred spruce
#

🙁 sadness ...

native eagle
#

@broken birchHello, as a friendly reminder we cannot generate images or videos in the server! Please do not post your prompts in the channels as it will not generate. If asking for advice or tips on your prompts do so in the appropriate channels.

still granite
obsidian fern
#

Evening gang. Just created an Opal that uses the LSE data, plus national and international politics to recommend the TYPE of stock you should buy/sell/avoid as well as Forex

Thought I'd give you guys a chance to use it

Just click start
Then in chat box type the date

wanton night
#

I created a Linkedin response generator. Step 1 - provide the link to the linkedin post Step 2- Optionally provide inputs on the style of response Step 3-Wait for the output and copy 1 of the 3 responses. https://opal.google/app/1XGPIwxIfsVXzPJCy8tErL4qSJ8Z7a8LD Please test and let me know your thoughts?

jolly dust
rich tartan
bold dome
#

I am in India using my Apple ID ending with iCloud.com through my iPad as Google ID which has Google pro subscription active. Previously when I used to open Gemini I used to be able to see Opal gems in the Gemini app but they have disappeared now and when I’m trying to directly login into opal I end up on a plain white blank page after logging in. This is stressful can someone help.

fallen jasper
#

Check it

#

Maybe the feature rolled out

bold dome
# fallen jasper Is it the same account with the pro subscription??

It was already there, the feature. But now it’s not. Like I said I only see the description but not the opal gems in the Gemini app. And when I’m trying to access them through opals website as soon as I login the page turns completely blind white. No matter how many times I refresh it stays the same. I deleted the cash cookies tried opening it in incognito and other browsers too. In fact I deleted the entire activity history from Gemini app as well. I don’t know what else to do.

fallen jasper
#

Maybe it's gotta do something with google servers

#

Maybe they took it away

hallow thorn
#

hi im from korea, recenly when I use opal, it always reached image generation qouta. is there any documents that i can check qouta of image, video generation? and how can I raise my limit?

sleek rampart
brave galleon
#

Hey, I have a question. I'm from Germany, and germans love their data protection, which is why we use Gemini Enterprise with EU servers in our company. Now to my question: is there a Gemini Enterprise API for integrating Gemini Enterprise into an N8n Workflow?

fallen jasper
#

I think yes

icy willow
#

@bold dome

floral barn
gaunt igloo
floral barn
#

What are the Top Opal usecases here?🙂 I have been very busy making a bunch and sharing them free with my friends. I would love to see what others have made here?😊

woven glen
floral barn
# zinc storm

What type of limits does opal have per day? Does it tell you when they are almost reached?

zinc storm
sleek rampart
#

What type of limits does opal have per

hybrid lake
floral barn
obsidian gust
#

Hi, I am generating an image and video using opal but I am facing the attached error. What does it mean? How do I know which part is the error? Also got the second error

obsidian gust
floral barn
outer bough
woven merlin
sleek rampart
#

Hi, I am generating an image and video

#

google_labs Hey Builders - In order to keep track of errors please post in opal general channel in the mean time we are looking to organize the channels to streamline better!

outer bough
fading marten
#

Hello

floral barn
#

Opal to prepare you going into a meeting, providing you with top questions and answers to ask and possible roadblocks/how to overcome them. - https://opal.google/app/1M6Gopeoc3m3Is-Ijpaoy8G4-W38moQjT (May be able to be remixed into preparing you for an exam or test. That will be my next one, will try & share.)

More of my opals - https://www.aiagentslibrary.com/go

#

Tip: Request it to include source URLs in your output, to allow you to trace the origin of its output if needed.

glacial hazel
#

Imagine a workflow where your raw footage transforms into a professional edit automatically. This isn't just theory; it is the next step for Opal.

The core concept is simple: you feed the agent your raw video, and it generates the perfect B-roll. It doesn't just create content; it acts as a director. It tells you exactly where to place those clips—Google VEO 3.1 generated assets—to maximize engagement. You drop a one-minute video, and the agent maps the sequence for you.

Currently, we face a technical bottleneck: duration. Most tools are locked into 8-second clips. That’s too rigid. For dynamic editing, we need a selector. Ideally, we could generate 3-second clips on command. Since the human eye needs a visual shift roughly every three seconds to stay engaged, this feature is non-negotiable. Without it, you are stuck doing manual, tedious cuts. Opal needs to bridge this gap.

To make this a reality, the System Prompt is your most powerful tool.

Don’t underestimate the configuration inside the generation node. If you aren't crafting that prompt carefully, you are leaving performance on the table. Use the "suggest an edit" feature—the magic pencil icon—to refine your instructions. Let the AI help you define the workflow.

glacial fox
wispy surge
#

Hi everyone! I’ve developed a specialized translation AI on Google Opal designed to handle long-form and complex technical documents.
What makes it different:

Adaptive Readability: It can bridge the gap between high-level research and general readers, making PhD-level papers understandable for high school students.

Cultural Localization: Instead of literal translation, it swaps idioms, proverbs, and cultural examples into natural equivalents in the target language.

Dynamic Table/Chart Translation: It specifically processes and translates data/content inside tables and charts.

Terminology Consistency: Ensures unified technical vocabulary across the entire document.

Smart Annotations: Automatically identifies and adds footnotes for niche concepts that readers might not recognize.
Link: https://opal.google/app/1vUgMeukj_rpDjZ26Rb1Rzt5lWAiYTTzA

edgy breach
serene yarrow
#

Medical Tweet & Quiz Expert 🩺
So I have built a Medical Tweet expert Opal that helps me to get cited and verfied texts and generate professional tweets from the response. Here is the description of the Opal.

  1. The Concept: A dual-mode AI assistant designed for medical professionals and educators to bridge the gap between complex research and social media engagement.

  2. How it Works: The app uses a dynamic decision-making workflow to handle two distinct medical tasks:
    i - Research-to-Social Pipeline: Drop a journal link or DOI, and the app scrapes the abstract to generate a 3-column Markdown table. It provides three unique "flavors" of tweets (Impact, Molecular Mechanism, and Clinical Outlook), automatically appending the full source URL for verified citations.
    ii - Multimodal Clinical Quiz Master: Upload a medical board question or an image (medical imaging or general imaging). The app utilizes vision and search grounding to provide a structured diagnostic breakdown, including the pathophysiology and a verified source handle.

  3. Key Features:
    i - Smart Extraction: Uses search grounding to bypass "Human Verification" walls on major journal sites.
    ii - Responsive UI: Renders outputs into a professional, "clinical-blue" themed webpage with distinct sections for tables and prose.
    iii - Precision Focused: Strictly adheres to medical accuracy and professional formatting—no AI fluff, just the science.

Take a look: https://opal.google/app/1N2Hw6MiZnIWvLAdqo7wnGNqCfdwCc2Fs

floral barn
#

Students Notes to Game Generator.🐸

What it does?
This Opal takes your text and instantly codes it into a fully functional, interactive web game that you can play right in your browser to learn.

Personalized for you
The best part? You get to choose the game type and theme to keep it personal and fun. I choose frogs, but you can choose "outer space with corgis or pizza!".😊

Nodes I used:
Drop your notes: Paste in your lecture, textbook summary, or meeting notes.
Pick your game type: Choose between a Trivia Showdown, Flippable Flashcards, Fill-in-the-Blank, or a Text Escape Room.
Pick your theme: Turn your Biology notes into a sci-fi space survival game, or your History notes into a retro 8-bit trivia battle!

Access it here:
https://opal.google/app/12vt4N4vm02uQn3M2IdB5jsRA39jfskdQ

floral barn
sleek rampart
#

google_labs Who remixed their first Opal this week? Show us what you made

rugged wave
#

hello, do you have any plans to develop opal mcp??

hazy pier
#

can anyone explain what the "code execution" tool does and how to use it?
I’ve tried searching, but I couldn’t find any documentation about it.
thanks for helping

tiny dust
#

Prompt:Ultra realistic 4K cinematic wide shot of a young Indian girl Simran (age 20, simple village

outer wyvern
#

Hello as a friendly reminder, videos and images are not generated within the server, you'll need to head to https://labs.google/, pick a tool, and generate it there. If you are asking for prompt advice please do so in the appropriate channel. Thank you!

sleek rampart
leaden fox
#

Anyway, I joined the discord because it shouldnt be getting this from anything in the prompt or input and is slightly concerning

#

@sleek rampart - If you're with google, could you help me look into this?

leaden fox
#

<@&1366831844502274180> - Any help with this? I use opal for small product demos and Cant risk this coming up in any instance.

outer wyvern
leaden fox
#

@outer wyvern Thank you, and thanks for deleting the image, I will send it to mod mail

outer wyvern
leaden fox
#

@outer wyvern Who does the mod mail go to and will I hear back about a resolution?

drowsy juniper
#

Opal

outer wyvern
sleek rampart
sleek rampart
sleek rampart
#

labs_dino Morning Chat! What are we cooking today? Also a friendly reminder please use ⁠labs-bugs-feedback・📥 to record your bugs if you don't have the channel you have two ways to add it:

Recommended: Double/right-click the ⁠Channels & Roles tab (top-left of the server) and select "Show all channels". Whenever a new channel opens in Labs, it'll automatically be added to your channel list!

You can also click ⁠Browse Channels and individually select the channels and categories you'd like to have visible in your channel list.

fallen jasper
#

btw can u add the tickets bot too

#

or is modmail enough?

sleek rampart
#

Rise and Shine ✨ Chat!

violet fern
violet fern
floral barn
#

I created a website in Google Canva that goes to a student friendly Opal podcast generator.

#

I will add the Opal link here for others once I am on my laptop. So grateful for your team's work on your Opal platform.

floral barn
hollow thorn
#

guys iam creating my personal agent in python, and i need opal extracted data ( it extract data from a youtube video ) to my python agent.
anyway to do so

plain sundial
#

Is there a way to convert a Gemini Gem into an Opal?

wide flare
#

Hi

dusky glade
#

How to app voice

north anchor
#

guys iam creating my personal agent in

fiery oracle
#

Opal gives an error when attaching a file… why is this happening?

north anchor
#

Opal gives an error when attaching a

north anchor
#

Is there a way to convert a Gemini Gem

sleek rampart
fiery oracle
#

Can it be saved to a local drive from the output?

north anchor
#

Can it be saved to a local drive from

sleek rampart
#

🚨 FYI: How Opal Sharing actually works (read before you share)

🔒 Default state
Opals are private by default, which means only you can see them until you decide otherwise.

🔗 When you share
You can either share with specific people or open it up to anyone with the link.
That part is simple — what matters more is how they can access it.

⚙️ Access settings (this is the important part)
There’s a toggle that controls editor view + remix, and this changes everything:

👀 Editor View ON
People can see your full setup — your graph, your prompts, and how everything connects.
They can’t edit your version directly, but they can remix it, meaning they can duplicate your work and build on top of it.

📱 Editor View OFF
People only see the app experience.
They can use it, but they won’t see your prompts, logic, or structure behind it.

⚠️ What most people miss
When you share an Opal, you are also sharing the linked drive file, which can sometimes create indirect ways for others to access more details than expected.

🧠 Best way to think about it
If you’re okay with people learning from or building on your work, sharing with editor view can be powerful.
If you’re trying to protect your workflow, prompts, or logic, you should be much more cautious.

🚫 If it’s sensitive
The safest move is simple: don’t share the Opal at all.

floral barn
plain sundial
sour island
#

the right click button isnt working in oppal i always use ctrl V to paste is this a problem or just waiting update to resolve it

north anchor
#

the right click button isnt working in

plain sundial
peak pine
vital oracle
wanton night
#

Trek Planner, built on Google Opal that acts as an elite wilderness guide and athletic trainer all in one.
You just input your target hike date, trail name, home zip code, and fitness level. The agent synthesizes trail distance, elevation, and 10-year weather averages, but my favorite feature is the "Local Practice Optimizer". It calculates the weeks until your hike and maps out a progressive training plan using local parks, hills, or urban stairs in your specific zip code to mimic the target mountain's elevation. https://opal.google/app/1ykyeJHyEBDpm8tc21FVVcv9T0aYLjV8y

fervent lance
north anchor
slow ginkgo
scarlet island
#

Hi everyone,

I am looking to build an agentic workflow to analyze SEC filings for equity valuation and investment risk management. How should I structure my prompts and instructions for Opal to achieve this?

cinder ravine
#

hey guys, I tried an Opal to fully automate my faceless YouTube Shorts of stoicism niche. Right now, it takes a single video idea and generates the script, the .mp3 voiceover, and the image prompts all in one click.

I'd love some raw feedback. If you try it, let me know what you think of the node structure or if you have any tips on making the outputs more consistent across different niches!

https://opal.google/app/1uOjdrwY7hCXAbCKcw1MvDijoVZ1TWJtI

spice rune
#

Hello , we wanted to remind you of our server guidelines, stated in #start-here・📍 : “Currently, our team can support engagement in English; the discussion in this server will largely remain in English for now.” Please keep in mind that non-English discussions and content may be removed. Also images and content cannot be generated in this server. Head on over to https://labs.google/ pick a tool and start generating. Thank you!

scarlet island
#

Hi everyone, Once Opal is set up, does it accept JSON and HTML files in the app?