Weekly Kit newsletter performance digest to Slack

Every Monday at 8am, post a Kit (ConvertKit) newsletter performance digest to a Slack channel so I can see how last week's sends actually landed. This is a deterministic pipeline (no reasoning required): list, fetch stats, format, post.

Trigger: cron, weekly on Monday at 8am (my local time zone).

Step 1: Call Kit's List Broadcasts and filter to broadcasts whose send time falls in the last 7 days (the current-week window). Also pull the prior 7-day window (days 8 through 14 ago) in the same call or a second call so we can compute a baseline. If no broadcasts were sent in the current window, skip to Step 4 and post a short 'no sends last week' note instead of an empty table.

Step 2: For each broadcast returned in the current window, call Kit's Get Broadcast Stats and collect: subject line, total recipients, open rate, click rate, and unsubscribes. Do the same for the prior-window broadcasts, but only keep their click rates so we can average them into a baseline.

Step 3: Compute the prior-week average click rate across all prior-window broadcasts. Then format the current-window broadcasts into a ranked table, sorted best-to-worst by click rate, with columns: subject, recipients, open rate, click rate, unsubscribes. Under each row (or as a trailing column) show a one-line delta vs. the prior-week baseline (for example: '+1.2 pts vs. last week's 3.4% avg' or '-0.8 pts vs. last week's 3.4% avg'). Format the table using Slack mrkdwn (backtick-wrapped code block is fine so columns align).

Step 4: Use Slack's Send a Message to post the digest to a channel I pick during setup. The message should have a short header line ('Kit newsletter digest — week of <date range>'), the ranked table, and a one-line footer with the total sends and total recipients across the week. Substitute the em dash in the header with an en dash or a colon if the platform prefers.

Integrations: convertkit (List Broadcasts, Get Broadcast Stats), slackbot (Send a Message). Prefer the slackbot integration over user-token Slack because this is an automated recurring post, not a message sent 'as me'.

Keep the pipeline strictly deterministic. No LLM step, no summarization, no editorializing. Just fetch, rank, format, post.