// the operator's workspace

Claude writes, thinks, analyzes.
robOS remembers and verifies.

What you decided, what you learned, what you promised a month ago. Everything is saved to files you can see, export and keep. And when it's not sure about something, it asks before writing.

Built by Adrian Ulmeanu, 30+ years in IT and automation. I use robOS every day for my own work.

"Search brings you an answer and forgets it. A wiki accumulates and grows."

Ady Insights, on the pattern robOS delivers ready to use.

// pick a scenario:

robOS terminal

// the problem

Claude remembers. But not the way you need.

Yes, Claude has had memory since March 2026. It writes a summary every 24 hours, saves what it decides to save, and you don't see exactly what was kept and what was left out. You can't separate one workspace from another inside its memory.

If you work across multiple workspaces, you end up repeating the same things in every new conversation. Not because it forgets entirely, but because it remembers without order.

robOS adds order, discipline and verification on top of what Claude already does well.

// what happens without structure

Without persistent memory, you repeat context from scratch in every new conversation. Yesterday you explained everything. Today you start over.

Without methods that learn from your feedback, every result sounds the same. What it writes for you is identical to what it writes for anyone else.

Without verification, you get answers that sound good but don't hold up. You publish them, then find out too late they were invented.

Without workspaces, when you switch clients you lose all the previous context. Tone, decisions, history. Everything has to be rebuilt manually.

just Claude

1

Open the old chat, scroll to find where you left off

2

Claude kept what it wanted, you don't know exactly what

3

All context is mixed into one long thread

4

Looking for a decision from 3 weeks ago? Good luck.

5

Switching clients? Re-explain everything from zero.

with robOS

1

Open it, see yesterday's summary in clean order

2

Each workspace has its own space, clean and separate

3

Find any old decision from two keywords

You control what gets kept and where

Same tools, but with order, memory and control.

// comparison

Without robOS vs With robOS

Same tasks, two ways of working.

Claude decides on its own what to keep and what to drop, without asking you

You see everything that's been saved, search through it, export it whenever you want

You lose what you built last month

Everything you've built accumulates, searchable in 2 seconds

Claude answers even when it's not sure about what it's saying

When it doesn't know, it asks

You re-explain each workspace context every time

Each workspace context loads automatically

Copy-paste between 8 open windows

One place for everything

Working method reinvented every time

Ready-made methods that learn from feedback

// what robOS adds

Three things Claude alone doesn't do.

Memory on 5 levels

Everything you worked on, decided, learned. On your disk, searchable from two keywords.

1

Daily memory

One file per day, with what you did, what you decided, what stayed unfinished. Clean format, readable in Notepad. context/memory/YYYY-MM-DD.md

2

Unclosed session detection

If you closed abruptly yesterday, today's first interaction tells you what's left hanging. You don't start from zero.

3

Decision journal

Important choices are written down with their reasoning. Six months later you know why you chose that way, instead of rebuilding from memory.

4

Per-method lessons

If a skill makes the same mistake twice and you correct it, the system remembers the correction. Next time it won't ask for the same thing. context/learnings.md

5

Audit history

How quality has evolved over time. You see the trend, not just today's state. context/audits/

Anti-improvisation verification

"I don't know" is a first-class answer, not a weakness.

Before writing anything important, it checks whether it can back up each claim

If it can't, it says "I don't know, I'll ask", it doesn't invent

When a skill is missing for what you asked, it says so directly. No workaround improvised.

"I don't know" is a first-class answer, not a weakness

robOS terminal

you:

Write an article about 2026 SEO trends with concrete numbers

robOS:

I found verifiable data for 4 out of 6 points. For the other 2 I don't have reliable sources, I'd rather skip them than invent. Want me to continue with the 4, or will you find the sources for the rest?

Multiple workspaces. Zero mess between them.

A workspace is your container: a client, a personal project, a partnership, a product line. Anything you'd want to keep separate. You switch with one command.

All context switches automatically: voice, brand, memory, output

Tomorrow you open the terminal, you're still in the same workspace

The price doesn't grow with the number of workspaces

"which client is active" see who you're working for now
"switch to Flavor Bistro" everything changes automatically
"client root" back to your base space
active client: Flavor Bistro

Voice: warm, personal, host tone

Audience: urban 25-35, foodies

Tasks: recipe article, Instagram post, newsletter

Last session: Tuesday, newsletter draft

active client: DentPro Clinic

Voice: professional, trustworthy, empathetic

Audience: patients 30-55, dental anxiety

Tasks: FAQ article, Google review response

Last session: Monday, competitor audit

One command. All context switches. Nothing to configure manually.

robOS workspace

$ switch to DentPro Clinic

Active workspace: DentPro Clinic

voice: professional, trustworthy, empathetic

audience: patients 30-55

last session: Monday, competitor audit

Workspace memory loaded. 3 open threads.

// your property

Your data stays on your computer. And stays yours.

Many AI assistants keep everything you discussed in their cloud. When you cancel the subscription, you lose it all. robOS works differently: everything that matters sits in files on your disk, in a format you can read with any editor.

~/robos/context/

Everything is on your disk

Memory, decisions, audits, brand voice, client profiles. Plain text files. Open them with Notepad, Word, any editor. Nothing encoded, nothing locked.

.env

API keys stay with you

They don't pass through any server of mine. Your Claude key, your Zernio key, any service you use - direct between you and that provider.

cp -r robos/ ~/backup/

Leave with everything, anytime

Moving to another computer? Copy the folder. Want an archive? Onto a USB stick. No account to close, no export to request, no cloud to migrate from.

// open source

The code is readable

You see exactly what robOS does, line by line. Nothing hidden behind a logo. If you want to change something, you can.

// a working day

What a day with robOS looks like

09:00

Open robOS, see what was left unfinished yesterday

09:15

Continue yesterday's task, without re-explaining context

11:00

Write copy for Flavor Bistro, with their voice auto-loaded

14:00

Switch to DentPro with one command

16:00

When the AI doesn't know a detail, it asks instead of improvising

18:00

Daily memory saves itself

// comparison table

Manual with Claude vs With robOS

Action Manual with Claude With robOS
Continuity between sessions Start from zero Pick up where you left off
Searching the history Scroll through old conversations Search your second brain, 2 seconds
Saving daily memory You have to tell it It saves itself, stops you from leaving without it
Learning from mistakes You repeat the same mistake The system learns and proposes a rule
Uncertain claims The AI answers anyway It tells you "I don't know, I'll ask"
Old decisions You forget them after 3 weeks The journal keeps them, you re-read 6 months later
Workspace context Re-explained every time Auto-loaded when you switch
Workspace switching Mental, complex One command, persists between sessions
Context when opening 10-15 min re-explaining who you are The system already knows, you start working
Working method You invent a flow for each task Ready-made methods that learn from your feedback
Audits and reports You run them manually, when you remember They run on their own, you see results in the morning
Large tasks You wait sequentially Work in parallel, 3-5x faster

// ownership

Plain files, on your laptop

  • > Notes and journals are files in a folder. Open them with any editor.
  • > Search runs locally. Nothing leaves for a vendor.
  • > Backup = copy folder.
  • > If you drop robOS, your data stays yours. Standard markdown files.

If you use Obsidian, open the robOS folder as a vault. You get instant graph view, backlinks, tags. If you don't, robOS works perfectly on its own.

// what you can do with robOS

27 skills across 5 categories

Pick a category, then a skill. See exactly what you get.

Give an input, get a published result.

YouTube to short clips

TikTok, Reels, Shorts, burned-in subtitles

Idea to social post + images

7 scenarios: text, video, theme, repurpose

YouTube to PDF ebook

Magazine style, feature article, not a summary

Topic to HTML presentation

Navigable deck, PDF export

robOS

you:

I have a 45-min YouTube video about 2026 marketing trends. Make short clips.

robOS:

Extracted 5 key moments from the transcript. Clip 1: "Retargeting is dead" (42s). Clip 2: "Email beats social" (38s). Reframe 9:16, burned-in subs, face tracking. Published on TikTok + Reels + Shorts.

Have your own methods? Add them in the same format. They integrate just like the included ones.

Some skills use external services: Zernio (publishing across 13 platforms), Apify (LinkedIn scraping), OpenAI or Gemini (AI images), HyperFrames (motion video), WhisperX (local transcription). Each has a free tier or low cost. Details and required keys in .env.example at install time.

// automations

Runs on its own, without you

Daily reports, monitors, validations. Scheduled to run on their own.

At 8am you see yesterday's report. On Mondays you get a review of what was learned last week.

You don't have to start anything. The system works in the background.

robOS does on its own
  • Saves the daily memory
  • Runs the morning audit
  • Generates the weekly report
  • Learns from repeated mistakes
  • Stops you if you try to leave without saving
you do
  • Decide what's important
  • Approve what gets published
  • Switch between workspaces whenever you want
  • Set direction, set priorities
  • Add new methods when you need them

You lead. robOS handles the routine.

4x

// productivity

You work faster because you stop redoing everything every time

Four concrete mechanisms that save you time.

context ready

The system already knows the voice, brand and past decisions for the active workspace. You stop wasting 10-15 minutes every time you open it, explaining who you are and what you do.

methods that learn

Repeatable flows that remember your voice, your feedback and what worked last time. Pick a method, tweak it, it learns. You stop inventing a flow from scratch.

automated routines

Daily audit, monitors, weekly lessons aggregator. They run overnight on their own. In the morning you see the result, you don't run them yourself.

parallel work

When the task is big (audit across 4 dimensions, research across 5 sources, repurposing across 8 platforms), the system puts multiple workers in parallel. You finish 3-5 times faster.

~/robos/context/memory/

$ ls -la

2026-05-12.md 2.4K today

2026-05-11.md 3.1K yesterday

2026-05-10.md 1.8K

2026-05-09.md 2.9K

$ cat 2026-05-12.md

## Session 1

### Goal

Refactoring landing page copy + deploy CF Pages

### Decisions

Moved from skill-count positioning to value positioning

### Open Threads

- Product screenshots: do before launch

Here's how it looks in practice: one command, a video built from scratch.

$ robOS run content-repurpose

Input: 45-min YouTube video

Extracting transcript... 5 key moments found

Reframe 9:16, subtitles, face tracking

5 clips generated. 38s, 42s, 45s, 51s, 33s

Published: TikTok + Reels + Shorts

// foundation

Built on a pattern the community has validated

In spring 2026, Andrej Karpathy described a way of working with AI he called "LLM Wiki": raw sources, AI-compiled pages and a set of rules that keep everything in order. Several practitioners tried to build it on their own. robOS starts from the same principle, but adds everything a solo user doesn't build for themselves: automatic verification, per-workspace memory, installable methods and routines that run without you.

schema how the AI works
wiki/ AI-compiled pages
raw/ original sources (PDFs, notes)

The idea is public. robOS ships it ready to work.

// what practitioners say

Independent conclusions

"That's not a productivity hack. That's a cognitive operating system."

Pilevar, Director of Analytics

"Search brings you an answer and forgets it. A wiki accumulates and grows."

Ady Insights

"Stop thinking of your notes as finished products. They are raw material."

Iakubov

Three expert practitioners, the same direction. robOS ships it ready to work.

// honesty

What robOS does NOT do

x

It does NOT replace Claude.

All the reasoning, writing, decisions. Claude does those.

x

It does NOT work without Claude Code.

If you don't have an Anthropic account, you don't have a product.

x

It is NOT an app with a login.

It runs on your laptop. Your data stays with you.

// full costs, no surprises

Total costs

Your Claude subscription

What you already pay. No increase.

robOS, $97 early-bird

One-time payment.

x

Dedicated server. NOT needed.

x

Other subscriptions. NO.

x

Monthly robOS fees. NO.

Total = what you already pay + $97 once = everything.

// install

Three steps, 15 minutes

1

Clone a folder

30 seconds.

2

Answer 5 questions

Guided onboarding, about 10 minutes.

3

Work normally with Claude

robOS is already there.

If you know how to talk to Claude, you know how to use robOS.

// is it for you?

Who it's for, who it's not for

Yes, if:

  • You work across multiple workspaces: clients, personal projects, partnerships
  • You already use Claude Code or Claude Desktop
  • Memory has to survive days, weeks, months
  • You have protocols to follow (audit, brand voice, validations)

No, if:

  • - You use Claude for quick questions
  • - You want a magic button without understanding anything
  • - You don't use Claude Code

// who built robOS

Adrian Ulmeanu

Adrian Ulmeanu

Founder, RoboMarketing

Creator of Agents Factory, a hands-on course where you build a complete AI agent system on your own server. I run robOS daily for my own projects and for clients.

"When you feel you need one more person, you probably need smart automations."

27 pre-built skills 8 dashboard sections 3 automated cron jobs

// community

You're not alone

  • > First month: included in the price
  • > Questions, examples, networking with other operators
  • > Direct support from me

// pricing

Price and access

Early Bird

97 USD

One-time payment, no subscription.

  • Everything described above
  • Operators group (first month included)
  • WhatsApp support
  • Free v1 updates
Get robOS, $97 early-bird

// frequently asked questions

Questions

I already have a Claude subscription. Why pay $97 on top?
Claude gives you conversations. robOS gives you structure: per-workspace memory, methods that learn, automations. Like the difference between an engine and a complete car.
How do I know robOS won't make things up?
For important claims, the system verifies before it writes. If it can't back something up, it tells you "I don't know this, ask me or check". And when a skill is missing for what you asked, it says so directly instead of improvising a workaround.
What if I forget to save my work?
Daily memory saves itself at key moments. If you insist on closing the terminal without saving, the system stops you. It won't let you lose your work.
What if I take a 3-day break?
When you return, it tells you "last session you were left with X, Y. Continue?" Without you digging through files.
Can I build this myself?
Yes, in 2-3 months. Pilevar built it over a weekend, but without hooks, without multi-workspace, without an update path. Or pay $97 and get the whole thing tested from day one.
Is it hard to install?
One command, 30 seconds. Then 15 minutes of onboarding. If you know Claude Code, you know robOS.
Does it work on Mac and Windows?
Yes, wherever Claude Code runs.
What if I don't use Claude Code?
Then robOS isn't useful to you for now. It's built on top of Claude Code.
Where does my data go?
Your data stays on your laptop. Memory, brand, output. The only external connection is Claude itself, through your own Anthropic account, as usual.
I only have one workspace. Is it worth it?
Yes. Memory, methods and automations help even with a single workspace. Multiple workspaces are a bonus, not a requirement.
Do I get updates?
v1 for life. Bug fixes free. New major versions sold separately.
What if I get stuck?
Write on WhatsApp or in the group. I help you with no time limit in the first month.
How fast do I see results?
Day one. You install, onboard, and work with the first method.
Do I need external accounts?
Memory, verification, workspaces and the dashboard work without anything extra. For skills that publish or generate (AI images, video, social media publishing, scraping), you need accounts with external services: Zernio for publishing, OpenAI or Gemini for images, Apify for LinkedIn. Each has a free tier or low cost. At install time you get a .env.example file with everything that needs configuring.

// next step

Ready to start?

📖 🪢 zzZ 👍 👋
Ask about robOS
Hi! Ask me anything about robOS: what's included, how much it costs, how it works. If I don't know, I'll point you to Adrian on WhatsApp.