Daily report — your stand-up, written from your own history
What you did yesterday, what follows from it today, and what got in your way — summarised from the record of what you actually asked your AI CLIs, by one of those CLIs, on your own machine.
How to do it
- Open My work. The report for yesterday is at the top of the page.
- Read the three sections: Yesterday, Today, Blockers.
- Press Copy and paste it into Slack, Teams or wherever your stand-up happens.
- Use the date picker beside it to read any of the last sixty days.
A day split across two tools comes back as one day
Morning in Claude Code on the API, afternoon in Codex on the migration. The report reads every CLI whose history LLMLinq can see, so it describes the day you had rather than one half of it.
A day is your midnight to midnight
00:00 to 23:59 on your machine's own clock, not a server's. Yesterday's report is prepared from 03:00 and shown from 08:00, again on your clock. If your machine was asleep at three, it prepares on the first check after it wakes.
Nothing about your day leaves your machine
The summary is written by whichever AI CLI is installed here, under your own login — the same tools the rest of LLMLinq drives. If none of them can answer, an Ollama you already run is asked instead. No hosted model of ours ever sees it.
It says what it read
Under every report is the CLI that wrote it, how many prompts it read, and which projects they came from. A summary you cannot check is a summary you cannot use, so the evidence travels with it.
Today is inferred, not invented
There is no record of tomorrow, so the plan is read out of what was left unfinished — threads still open at the end of the day, work clearly mid-flight. Where the record suggests nothing, it says so rather than filling the space.
The first morning
Nothing is prepared before the first overnight run, so the panel offers to write yesterday's report immediately instead of showing you an empty box.