TLDRs

Read a briefing at a glance

Home shows your newest unread TLDR with a short summary. Choose Read briefing to expand the recap and its source counts, or All TLDRs for older briefings. The history also starts with summaries, so you can scan several periods quickly.

New briefings are written for a quick read: a one-sentence summary and up to three short bullets, prioritizing blockers, decisions, and meaningful changes. Quiet periods need only a sentence. Existing briefings keep their full text behind Read briefing.

Choose Questions to see the questions attached to a briefing, then select a question to reveal its answer, Suggested actions, and discussion. Switching between the briefing and its questions shows one at a time. Collapsing a section keeps an unfinished reply and your draft in place while you stay on the page.

Set up TLDRs

Open TLDRs from the company section menu. Owners and admins can configure the schedule; every Member can read the generated history. For a new company with no saved TLDR settings, hiring the first AI Employee starts an enabled Daily schedule automatically.

• That first AI Employee is selected as the writer. Connect an active AI Model before the first TLDR is due.
• Pick Every 4 hours, Every 8 hours, Every 12 hours, Daily, or Weekly. Daily is selected by default.
• Choose another writer or cadence, or pause automatic TLDRs, then save your changes.

Owners and admins can change the employee or cadence, pause the schedule, or choose Generate now when a recap is useful before the next interval.

Questions every briefing answers

A recap tells you what happened. The questions you have afterwards are usually the same ones every week — so write them down once instead of asking them every time. Under Questions to answer on the TLDR settings page, add up to eight standing questions.

• Add your own, or start from a suggestion such as What should we stop doing?
• Reorder them with the arrows — cards appear under the briefing in this order — and save your changes with the rest of the schedule.
• Switch one off to keep its wording without answering it for a while. Delete it to remove it for good.

As soon as a briefing is posted, the writing AI Employee works through the list and adds each answer under Questions. Open a question when you want its answer; you do not need to ask it again.

One-click actions on an answer

Where an answer names something concrete — "stop the nightly scrape, it has found nothing for three weeks" — the AI Employee attaches a button for it. Agreeing costs a click instead of typing the proposal back in your own words.

• Each button carries a short label and the full sentence of what pressing it will ask for. You see that sentence before anything runs, and confirming is what sends it.
• The AI Employee then carries it out with your access, not its own, and reports back on the same card. A finished button turns into a tick.
• Discuss opens the card's conversation instead, for anything a button does not cover.
• Not worth doing? Clear a suggestion with the × beside it. The card still records that it was suggested.

What a TLDR includes

Each TLDR covers one bounded period. It can summarize messages from public Workspace channels, company-visible journal entries, and terminal Routine Run output.

If that period contains no eligible activity, Genosyn creates nothing. Home and the TLDR history stay free of empty recaps.

Ask your own question about a TLDR

You can paste screenshots into the AI message box, drag files onto it, or use the paperclip. Review and remove attachments before sending. See image formats and limits.

Standing questions cover what you always want to know. For anything else, ask on the spot: choose Ask a question on a briefing. If it already has questions, open Questions, then choose Ask a question.

• Click a suggested question such as What can be improved?, or type your own.
• Each question becomes its own card, answered by the AI Employee who wrote the briefing, with its own buttons. The brief itself is never edited.
• Cards are company-wide, like the briefing. Anyone in the company sees them; the Member who asked, and any owner or admin, can remove one. Removing a card produced by a standing question leaves the standing question itself in place.

The first answer on a card is discussion-only, however the card was created. Genosyn runs it on the same restricted path the briefing itself uses, with no tools at all, and treats the recap as untrusted reference data — so text inside a briefing cannot trigger an action.

Discuss a TLDR without leaving the page

Reply on any card to keep talking. Follow-ups run as ordinary AI Employee Chat with your own access, on the same page as the briefing — so when the employee proposes something, you can simply ask for it. Add a routine for that creates the Routine, rather than another paragraph describing one.

A card and its conversation are company-wide and stay attached to the briefing. If the writer has since been removed, existing cards remain readable but no new question can be asked about that TLDR.

The writer can summarize, not act

Genosyn treats the source material as untrusted and runs the chosen employee's model in a restricted summarization turn. It receives no coding, browser, Integration, Genosyn, or company MCP tools. Writing a TLDR cannot send a message, change a record, or use the employee's Grants. The first answer on a question card runs on that same restricted path.

Working out which buttons an answer deserves is a second restricted turn with the same shape: it can submit a list of suggestions and do nothing else. Suggesting an action is mechanically incapable of taking one.

Acting is a separate step you take deliberately. Only a follow-up you send on a card, or a button you press after reading what it will do, carries tools — and it carries exactly your own access, never more, and never on the strength of something written inside a briefing.

Read and dismiss

The newest TLDR you have not dismissed appears on Home. Open TLDRs to read the complete generated history at any time.

Dismiss is personal. It removes the recap from your Home, but does not delete it or mark it as read for anyone else. Generated history is preserved.