Weekly competitor social listening briefing in Notion

Build me a weekly competitive intelligence agent workflow. It should run on a cron trigger every Monday at 7:00 AM America/New_York.

The workflow tracks a configured list of competitor brands. For each competitor, query Infegy Starscape over the past 7 days and call Get Topics Analysis to surface emerging themes with size and sentiment, Download Post Records (capped at a sensible sample, for example the top 500 posts ranked by engagement or reach so the run stays within plan and cost limits), and Generate AI Summary to produce a narrative read on the conversation. From these results, capture for each competitor: share of voice (this competitor's post volume relative to the total across all tracked competitors), week-over-week sentiment trend versus the prior week's pull, top emerging themes, and 3 to 5 standout posts (highest reach or most engaged) with author and link.

After all per-competitor pulls are done, synthesize a single comparative weekly briefing covering: share-of-voice ranking across competitors, week-over-week sentiment movement per competitor (call out the biggest mover up and the biggest mover down), emerging themes per brand with a short narrative on what is driving each, and any noteworthy content, launches, or campaigns competitors ran in the week.

Use Notion's Create a Page operation to drop the briefing as a new dated page in our designated competitive intelligence Notion database. The page title should be something like 'Competitive Intel — Week of [Monday date]'. Fill in the database properties (date, week number, competitors covered) and put the full briefing in the page body with clear section headings.

Once the Notion page exists, use Slack's Send a Message operation to post a short message in our marketing channel with a one-line summary (top mover for the week) and a link to the new Notion page so the team is pinged when the report lands.

The competitor list, the Notion database id, and the Slack channel should be configurable inputs so we can edit them later without rewriting the workflow. Make the post-record sample size and the AI summary tone (for example, neutral analyst vs. punchy marketing) configurable too. Store last week's per-competitor metrics so the next run can compute the week-over-week comparison.