⚡ Activation

💭 in your head · nothing to type

Catch a recurring information flow you run every week — one that lives in you now almost like a reflex.

Don't picture it as a routine chore. Picture it as your circulatory system: the heart (trigger) contracts and pumps blood in a loop. The lungs (processing) oxygenate it. The capillaries (delivery) carry the result all the way to every cell.

Run that loop in your head: where is the heart of your flow — what makes it contract? And where does the blood finally land?

👁 Reflection

💭 in your head · nothing to type

Picture a conductor opening a score and seeing it whole: first violins know their entry, brass knows theirs, percussion knows theirs. No one waits for a tap on the shoulder.

Look at your weekly flow the same way: which "voices" in it already know their part — things you do without thinking? And which instrument still needs you standing there, baton raised, to bring it in on cue?

That's the section where a pipeline belongs.

💡 Concept

Three pipeline types

1. Manual pipeline (today)

You run it manually when needed. Ideal for irregular tasks.

You → [URL/file] → Claude → result

2. Scheduled pipeline

Runs on a schedule. Good for regular tasks.

Every morning at 8:00 → collect RSS → analyze → report to Slack

There are three routes to a scheduled run: an operating-system timer, CI, or a no-code service. The comparison is in the table below. Claude Code Hooks don't belong here: they fire on agent events, not on a clock.