Opal upgrade is here

Hi ,

Google just added agent capabilities to Opal.

Instead of manually picking models for each step, you now select an agent that figures out which tools it needs. Memory, dynamic routing, and interactive chat built in.

Turns static workflows into interactive experiences.

Available in Google Labs now.

Today's website brand audit prompt and this week's community wins first. After that, what these agentic upgrades mean for workflow building.

🔥 Prompt of the Day 🔥

Website Brand Audit: Use ChatGPT or Claude

You are an expert UX and brand consistency auditor.

I need you to audit my website against my brand standards.

Essential Details:

• Company: [YOUR COMPANY NAME]

• Website URL: [YOUR WEBSITE URL]

• Pages to audit: [HOME, ABOUT, SERVICES, CONTACT — OR "FULL SITE"]

• Target audience: [WHO THIS IS FOR]

• Brand voice: [YOUR DEFINED VOICE/TONE]

Audit these six dimensions (rate each 1-5):

1. Voice & Tone Consistency — Does copy sound like same brand on every page? Any tonal shifts or personality contradictions?

2. Messaging Clarity — Is primary message immediately clear? Do secondary messages support without competing?

3. Visual Cohesion — Are colors, fonts, imagery, spacing consistent? Do visual choices reinforce brand personality?

4. Audience Alignment — Would target audience immediately understand who this is for, what problem is solved, and what to do next?

5. Trust & Credibility — Are social proof, credentials, guarantees present and appropriately placed?

6. Call-to-Action Consistency — Are CTAs clear, consistent in language, strategically placed?

For each dimension provide:

• Rating (1-5)

• What's working

• What's broken or inconsistent

• Specific examples with page/section references

End with:

• Overall Brand Consistency Score (1-5)

• Two-paragraph executive summary of biggest strengths and most critical gaps

🎉 Win Wednesday 🎉

This week's wins from the community.

Charles Janoah Policarpio shared his Claude-powered ad creation system.

His words: "I use Claude & a proven process to launch 3-5 decent ad angles in a day. I feel superhuman to start any business now lol"

His process:

Deep Research on three documents: Market research, competitor research, client/own product research.

Deep dive call with client to extract USP and Big Idea.

Get reference creative brief, break down structure, ask Claude to generate 3 angles in same brief format.

Use that brief for Claude to write ad using one of those angles.

Result: 3-5 ad angles per day.

Why this matters:

Most marketers spend days developing single ad angle. Charles ships 3-5 in one day.

He's not winging it. He's following proven process. Then using Claude to execute at scale.

Research feeds the brief. Brief feeds Claude. Claude outputs ads.

System beats speed. Process beats prompting.

Nancy Ruiz learned how to use Whisk.

Simple win. Valuable tool added to toolkit.

Sondra Verva created, sold, and taught 2-hour workshop.

Her words: "I created, sold and taught a 2 hr workshop, called Faceless YouTube Launchpad using the most INSANE frameworks that I turned into a teachable skill within Manus to help users launch everything needed for a YouTube channel that's fully researched & search engine optimized! Including scripts, video descriptions, keywords, and thumbnails!!"

What she built:

Frameworks in Manus for launching YouTube channels. Fully researched and SEO-optimized. Scripts, video descriptions, keywords, thumbnails included.

Showed how to use AI to create 2-minute videos in minutes with no editing.

Vibe-coded an app with everything students needed to follow the process.

Used Agentic AI to validate original work within YouTube itself.

Planning upsells to expand the offer.

Why this matters:

She didn't just create content. She created teachable system.

She didn't just teach theory. She built tools students use immediately.

She didn't just launch. She validated with AI research and planned upsells.

Product, workshop, app, validation, expansion plan. All executed.

What did you ship this week?

New system? First sale? Tool mastered?

Reply and share what you built.

Did You Know?

Public opinion on AI varies dramatically by country — strong majorities in China, Indonesia, and Thailand view AI as more beneficial than harmful, while optimism remains far lower in Canada, the United States, and the Netherlands, though sentiment is shifting positively worldwide.

🗞️ Breaking AI News 🗞️

Google Upgrades Opal to Agentic Workflows

Google announced major upgrade to Opal, transforming static workflows into agentic intelligence.

Available now in Google Labs.

What Changed

Previous Opal: Static model calls. Manually pick model. Predefine steps. Rigid formats.

New Opal: Agentic intelligence. Select agent in generate step. Agent determines best path. Triggers right tools and models automatically.

Shift from rigid configuration to dynamic execution.

The Agent Step

New agent step understands your objective. Figures out which tools and models it needs.

Triggers Web Search for research. Triggers Veo for video. Automates complex tasks with less manual configuration.

Proactively determines best path based on goal.

Example: Visual Storyteller

Before: Create storybook by predefining page counts and user questions. Static format.

Now: Visual Storyteller Opal where agent autonomously decides which details it needs. Suggests plot points. Lets you direct where story goes.

Dynamic, unique narratives shaped by real-time creative decisions.

Example: Interior Design

Before: Simple one-way process. Upload picture. Input style. Receive redesigned image.

Now: Interactive Room Styler. Feels like collaborating with design partner.

Upload photo of empty living room. Describe mid-century modern vision. Agent generates initial concept with era-specific decor and palettes.

Not right? Provide feedback on specific elements. Agent refines grasp of aesthetic. Researches niche design sub-styles. Creates redesigned images that feel uniquely yours.

Agent understands goal, thinks about best way to get it done, reaches out when it needs input, recruits best models and tools.

Three New Capabilities

Memory: Opals remember information across sessions. User's name, style preferences, running shopping list.

Opals grow smarter and feel more personal with use.

Example: Video Hooks Brainstormer stores user's brand identity and preferences. Generate tailored video ideas without repeating preferences.

Dynamic routing: Define multiple paths agent can follow based on custom logic. Describe criteria. Agent intelligently transitions to correct step when conditions met.

Example: Executive Briefing Opal tailors briefing based on whether meeting with existing or new client. Searches web for new client backgrounds or reviews internal meeting notes.

Interactive chat: Agent initiates chat with user to gather missing information or offer choices before moving to next stage.

Example: Room Styler Opal. If user didn't give enough detail at first, Opal keeps asking questions or showing examples.

The Philosophy

Best of both worlds: Power of AI agent working toward goal plus control of step-by-step workflow you can customize anytime.

Kept Opal simple while making it significantly more capable.

New users: Opals "just work" because agent is smart enough to self-correct, remember, optimize.

Power users: Standard fixed steps available whenever you need high-precision prototyping or rigid logic.

Bridging gap between automation and control.

Why This Matters

This is shift from configuration to conversation.

You don't tell Opal every step. You tell it what you want. It figures out how to get there.

Memory means context persists across sessions. No repeating yourself.

Dynamic routing means workflows adapt based on conditions. Not rigid paths.

Interactive chat means agent can ask clarifying questions. Not just execute blindly.

For users, Opal becomes more like collaborator than tool.

For developers, building workflows becomes faster because agent handles complexity.

What This Means

If you use Opal, workflows just got smarter and more interactive.

If you build workflows, less manual configuration required. Agent figures out tools and models needed.

If you create products, agentic layer makes experiences feel personalized and adaptive rather than static.

This is Google's answer to AI agents. Not separate agent product. Agent capabilities built into workflow tool.

Over to You...

What workflow would you automate if an agent could figure out the tools and steps for you?

Hit reply and tell me your first use case.

To intelligent workflows,

Jeff J. Hunter 
Founder, AI Persona Method | TheTip.ai

NEW: “AI Money Group” to Learn to Make Money with AI