Ask for the outcome.
Get the files.
Describe what you want in a sentence. LumenAgent plans it, opens the real editors, does the work, and hands back documents you can keep editing — while you watch every step or go and do something else.
One sentence in. The plan is a panel you can read, change or stop at; anything consequential waits for you; and what comes back is two files sitting in your library, already open in the app that made them.
Then it keeps going without you.
The plan and the approvals are what make one run trustworthy. These three are what make the next hundred worth having.
Every claim arrives with somewhere you can go and check it.
Research turns cite what they read — web pages you can open and documents named exactly — so you can check the number before you put it in front of anyone.
- Numbered, openable sources: your own files and the pages it read
- It names what it could not verify instead of averaging over the gap
- A long run leaves a trail: which app holds each result and what it came out of
Ask from your pocket. Walk back to finished work.
Anything that needs your real files becomes a job your own desktop runs — behind the same approvals, on the same local files. The phone starts it and tells you when it lands.
- The phone is sent an opaque job ID, never a path or a filename
- Your machine opens the editors and does the work behind your approvals
- Watch it or cancel it from the phone; the files are on your desk when you get back
Ask once. Then stop asking.
Any run worth repeating becomes an automation — on a schedule, when a file lands, when a client writes in, or fifteen minutes before a meeting starts.
- Watching costs nothing; credits are spent only when a trigger starts a run
- Rows marked Free replay a compiled recipe on your device with no model call
- Cap what one run may spend — at the ceiling it stops and reports
It does the work, and shows you all of it.
A chat assistant is a very good writer that cannot open your spreadsheet. The gap between those two things is where your afternoon goes.
Nothing happens invisibly
Understand, plan, work, deliver — written before anything is touched. The panel carries the checklist, what each phase actually did, the step count and the credits spent so far. A run that fails stops at the step that failed and names it.
It asks before it changes anything that matters
Reading your own library is routine. Renaming, moving, overwriting, deleting, emailing or publishing stops and waits — whatever permission preset you are running. You approve the exact list, decline it, or grant that one permission standing.
A checkpoint before every AI edit
The document’s previous state is saved before each revision, so an unwanted change is one click from undone. You also cap what a run may spend and how many steps it may take — at the ceiling it stops and reports.
It drives LumenPDF, LumenSheets, LumenDocs, LumenDesign, LumenSlides, LumenNotes, LumenForms, LumenApps and LumenFlow. Ordinary editing happens on your disk; content leaves the device only for the cloud step you asked for.
You never pick the app. It does.
Choose any of these and watch which editors and services the job runs through — and what you are left holding at the end.
LumenAgent · common questions
Will it change my files without asking?
No. Reading your own library is routine, but anything consequential — renaming, reorganizing, overwriting, deleting, sending an email, publishing a link — stops and waits for your approval, whatever permission preset you're using. Every AI revision also saves a checkpoint of the state it started from, so you can put a document back exactly as it was.
Do my files get uploaded?
Ordinary editing happens on your own disk. Content leaves your device only for the cloud step you asked for — an AI action, a share, a connector call — and only the content that step needs. Local files stay local unless you choose otherwise.
What does a run cost?
Agent runs draw on the same monthly AI allowance as the rest of LumenQube. Every account gets a free allowance each month, and you can cap what a single run is allowed to spend. When a run reaches your budget it stops and reports instead of continuing.
Which engine runs it?
Your plan decides how deeply Agent reasons. Short, mechanical jobs finish on Lumen Fast; long autonomous runs across many files are what Lumen Expert and Ultra are for. See what each engine does in every tool →
Can it use my Google account?
Only if you connect it, one service at a time. Drive, Gmail, Calendar and YouTube are approved separately, read and write permissions stay apart, sends and publishing remain reviewable, and you can disconnect from LumenQube or revoke from your Google Account at any time.
Can I use my own tools?
Yes — connect a Model Context Protocol (MCP) server and its tools become available in the same run, used alongside your library, the web and the LumenQube editors.
Can I start a job from my phone?
Yes. LumenQube runs a quick cloud Agent chat on its own, or hands the turn to your paired desktop — which has your files — and streams the result back into the same conversation. You can also dictate a goal instead of typing it, and have the answer read back.
What happens if a step fails?
The run stops at that step rather than guessing past it. You can retry that single step, change the plan, or cancel — the work already completed stays. An interrupted session comes back where it was.
Give it the outcome. Keep the files.
Free to start, free to use. Every account starts with a monthly AI allowance.
No install needed for the browser version — same account, same conversations.