Weekly HubSpot pipeline health report in Slack

Every Monday at 9am, post a sales pipeline health report to our team's Slack channel. This is an agent workflow because picking the top at-risk deals and writing the weekly commentary needs judgement on top of the raw numbers.

Pull our open deals from HubSpot using Search Deals, filtered to deals in our active sales pipelines and excluding anything already in closed-won or closed-lost stages. Include each deal's name, owner, amount, current stage, expected close date, and last activity date. Also call List Pipelines and Get Pipeline so the report uses real stage names and stage order, not internal IDs, and so the analysis can read each stage's probability for weighted pipeline value.

Spin up a fresh Daytona sandbox with Create Sandbox, then use Run Code in Sandbox to execute a Python script (pandas plus matplotlib) that computes: weighted pipeline value by stage (amount times stage probability), average days each deal has spent in its current stage, stuck deals with no activity in the last 14 days, and the week-over-week change in total pipeline value (compare to last week's snapshot, and persist this week's numbers so next week has a baseline). Render a single PNG chart of pipeline value by stage. Pull the chart out with Download File, then call Stop Sandbox so it does not sit on quota.

Post the report to our Slack sales channel as one message using Slack Bot's Send a Message, with the chart attached via Upload a File. Keep the Slack write-up tight so it reads cleanly on mobile: a one-line headline, the week-over-week pipeline value delta, the top five at-risk deals (name, owner, amount, days since last activity), and one or two sentences of commentary on what changed versus last week.

Make the Slack channel, the time of day, the at-risk threshold (default 14 days), and the pipelines or owners to include easy to change later. If HubSpot returns no open deals, post a short "no open deals to report this week" message instead of an empty chart.