The Goldie Knowledge Studio™
FREE · NotebookLM + Hermes video agent · one dashboard
The new NotebookLM upgrade is insane. And it's free.
NotebookLM creates. The Hermes video agent edits. Drop any URL, PDF, or transcript. Get audio overviews, slide decks, mind maps, infographics, and full videos in minutes. Then pull any of them into the new video forge — locally, free — to remix, trim, add your AI avatar, and ship them as finished social content. The bridge upgrade most people don't know exists, sitting one tab away in Agent OS.
"Your NotebookLM notebooks, audio overviews and chats — all in one place, synced to Obsidian. Generate audio, video, slide deck, mind map, infographic, flashcards, quiz, data table, report. Returns immediately, generation continues in NotebookLM."
— Agent OS · Mission Control · Notebook panel
II · my story · why this matters
I was you. Then I built this.
Before
Turning research into content was a nightmare.
I'd open NotebookLM, paste sources, hit generate, and get an audio overview I loved.
Then it just... sat there.
To turn it into video I had to open three more tools — one to transcribe, one to render visuals, one to add my face.
Each handoff lost something. The audio went one direction. The video another. The infographic ended up in a Downloads folder I'd never find again.
Nothing flowed end-to-end.
Then I wired NotebookLM into Agent OS — next to the new video forge.
After
Now I drop one source into NotebookLM.
Audio overview, slide deck, mind map, infographic — all generated, all saved.
Every asset shows up in the Notebook tab automatically.
I click any of them and pull it into the video forge — trim it, add my avatar, render the finished cut.
One source. Five finished pieces of social content.
Monthly tooling bill: $0.
You can have this too. Same notebook. Same forge. Same dashboard.
Before you scroll on —
Commit to transitioning today. Not tomorrow.
You've seen the proof above. Real people. Real finished content. Already being shipped.
The next 10 minutes show exactly how I wired NotebookLM into Agent OS next to the video forge.
So here's the deal.
If you're reading this — promise yourself one thing right now. You're going to finish this guide AND run one source through the Knowledge Studio before you sleep tonight. Just one. Because the moment you make this transition, your whole content workflow changes.
The people sitting still are getting passed. The people shipping today are the ones who'll look back in six months and say "that was the moment."
Be one of those people.
Commit to the transition. Commit to running one source through it today. This changes everything about how you make content from research.
IV · the framework
The Goldie Knowledge Studio™.
Five passes that turn raw knowledge into editable, ready-to-ship social content.
Each pass stands alone — but together they're a pipeline. One source becomes five formats. Each format becomes ten finished cuts. Each cut becomes ready-to-post content. By Friday you'll have shipped more than most operators ship in a month.
The five passes of the studio — Capture, Generate, Save, Edit, Ship.
i.
Capture — any source becomes raw material.
Drop URLs, PDFs, transcripts, web articles, your own old content. The Library inside the Notebook tab accepts whatever you give it. You stop hunting for inputs. The system reads anything.
ii.
Generate — nine artefact types from one source.
Audio overview, video, slide deck, mind map, infographic, flashcards, quiz, data table, report. Hit Generate. Walk away. Each artefact lands in the notebook automatically. You stop manually re-explaining the same idea in five different formats.
iii.
Save — every asset auto-synced to your vault.
Every generated artefact downloads to AGENTIC OS/notebooks/_assets/ and syncs to your Obsidian vault. Audio plays inline. Video plays inline. PNGs show inline. Click any of them in the Assets sub-tab and it's right there. You stop losing files in random Downloads folders.
iv.
Edit — pull any asset into the video forge.
This is the bridge that's new. Any video, audio, or infographic NotebookLM generates can be pulled into the Hermes video agent. Trim it. Add your AI avatar overlay. Layer in titles. Render the final cut locally — $0 per pass. You stop generating one-shot artefacts. You start editing them into finished work.
v.
Ship — finished social content out the other side.
The edited video lands in your Workspace tab next to every other render. The audio cut sits in the audio bucket. The infographic in images. All click-to-preview, all ready to post. You stop ending at "I generated something." You start ending at "I shipped something."
V · the bridge
The upgrade most people miss.
This is the part that matters.
NotebookLM has always been good at generating one-shot artefacts. The new upgrade — having it sit next to the Hermes video agent inside Agent OS — is what changes the workflow.
Before, your audio overview lived in NotebookLM forever. You'd play it once and forget it.
Now it auto-syncs to your vault. Then you pull it into the video forge with two clicks. The video forge accepts the audio as a source track. You add your avatar on top. You layer in the lower-third titles. You render the cut.
Same story for the infographic — pull it in as a visual asset, add motion, ship the finished animated explainer.
Same for the video — re-cut it, add your avatar overlay, change the music, render a different aspect ratio.
NotebookLM creates the raw.
The video forge turns the raw into the shippable.
Thinking it?
"I can already use NotebookLM directly — why do I need the bridge?"
Because the bridge is what turns generation into shipping.
Using NotebookLM alone, you generate an artefact and it sits there. You'd have to manually export, manually find the file, manually open another tool to edit it, manually upload to where you publish.
With the bridge, every artefact lands in your vault. The video forge can read from the vault. The output lands in Workspace. Everything is one tab away.
The bridge collapses five tools into one screen. That's where the time savings come from.
✓
Members report shipping more finished content in week one than they did in the previous month.
XII · why inside agent os
Why this can't live on its own.
NotebookLM alone is useful. The video forge alone is useful. But put them in the same dashboard, with shared memory and one-click handoffs, and the workflow changes completely.
a.
Shared vault across every agent.
NotebookLM saves to AGENTIC OS/notebooks/_assets/. The video forge reads from the same vault. So does Hermes, OpenClaw, and Claude — all of them can ground in the same knowledge.
You stop having to manually move assets between tools. The vault is the handoff.
b.
One dashboard, one tab away.
Notebook sits next to Video. Video sits next to Studio. Studio sits next to Goals. The whole content pipeline runs on one screen.
You're never more than two clicks from any stage of the workflow.
You stop managing five browser tabs. You start running one studio.
c.
Outputs compound across the stack.
Today's audio overview seeds tomorrow's video cut. This week's infographic becomes next month's animated explainer. Every asset is reusable in every other workflow.
One source isn't one piece of content. It's a seed every agent draws from.
d.
The bill stays zero.
NotebookLM is free. The Hermes video agent renders locally — free. Your own Obsidian vault is free. Even Free Claude Code (for fallback text) is free.
The whole pipeline costs $0 per month after setup.
The dashboard is the cost-killer. Pieces in isolation would leak SaaS subs.
Thinking it?
"I'll just use NotebookLM and a separate video editor."
You can. And in two weeks you'll be paying for both, wiring nothing between them.
The bridge is what makes NotebookLM go from "thing I generate in" to "engine that feeds my whole content pipeline."
Without the bridge, every artefact is a dead end. With the bridge, every artefact becomes raw material for the next.
You stop generating. You start running a content engine.
✓
Members who tried NotebookLM standalone first all moved to the full Knowledge Studio inside two weeks.
XIII · the voice in your head
Three beliefs holding you back.
✕ "NotebookLM is too basic to build a real content pipeline on."
It's basic on its own. Wired into Agent OS next to the video forge, it's the input layer for an entire content factory.
✓ NotebookLM is the engine. The dashboard is the chassis.
Same engine, completely different vehicle once it's plugged in.
✕ "I'll never use nine different artefact types — three are enough."
That's fine. The point of the nine isn't to use all of them on every source. It's that the right artefact is one tap away when the moment comes.
✓ Different sources need different formats — and now you have all nine.
For one source maybe audio + infographic. For another maybe slide deck + report. The studio doesn't force you. It offers.
✕ "AI-generated content all sounds the same."
Generic prompts produce generic content. Sourced prompts with your focus prompt + your specific angle produce content that sounds like you, with your data, in your voice.
✓ The studio amplifies your voice — it doesn't replace it.
Your source. Your focus prompt. Your edit pass in the video forge. The output is yours, just generated faster.
Don't take my word for it
258 real members already broke through these exact beliefs. Their wins — real notebooks, real videos, real outcomes — are documented here.
Read the 158-page testimonials doc →
