Skip to main content
You find out about at-risk deals too late. The CRM says they’re on track until they aren’t. Reps mention concerns in standup — for the deals they remember. The ones that go quiet across email, Slack, and calls slip out of forecast without anyone noticing. By the time the miss shows up in pipeline review, the window to save them is gone. Pipeline Risk Detection scans every open opportunity continuously and flags the deals that are silent across every channel — pulled from every source, delivered where your team works.

​
Data Sources

Starter (required only): a useful risk view based on CRM state and call engagement — deals with low recent activity, missed touchpoints, or stalled stages. Full (required + recommended): the same view, plus important context such as silent email threads, calendar absence, and Slack flags from your team. Same workflow — significantly more signal. See the Customer Context Graph for how each source feeds the brief.

​
Choose your path

​
Build in Clearskies

For teams whose daily surface is Slack. Build a scheduled risk scan once and your leadership team gets a prioritized digest of at-risk deals every Monday — delivered to Slack, no manual pipeline review.

​
The Workflow

In the Clearskies app, open the workflow builder and describe what you want in plain language:
The app configures the trigger (scheduled scan — weekly or daily), the data pulled (your connected sources), and the destination (a leadership channel, per-manager DMs, or per-rep DMs with manager cc). Review and refine before deploying.

​
Deploy to Slack

Once deployed, your leadership team sees this in the pipeline-risk channel every Monday morning:

​
Customize the rollout

  • Adjust risk signals to your team’s playbook (no economic buyer, no champion, stage-velocity drift, single-threaded deal)
  • Choose recipients (leadership channel, per-manager DM, per-rep DM with manager cc)
  • Set cadence (weekly digest, daily scan)
  • Tailor the brief format to your team’s standard (deal-focused, rep-performance, segment-by-segment)

​
Build in Claude or ChatGPT

For teams whose daily surface is Claude or ChatGPT — sales leaders working in either, or RevOps testing the workflow before deploying it to the team. Output appears in your conversation.

​
The Prompt

Open Claude or ChatGPT with the Clearskies MCP connected and paste this (or modify to your needs).
Claude or ChatGPT returns a digest structured like the Slack example above, generated from your data and formatted as a chat response. Refine the prompt to fit how your team works the pipeline.

​
Make it yours

  • Add Focus only on deals closing this quarter to narrow the view to what’s currently in forecast.
  • Add Show me which reps have the most at-risk deals to spot a coaching pattern.
  • Add Compare what reps are reporting in CRM to what the signals show to surface optimism bias.

​
The Skill (Claude)

The clearskies:pipeline-review skill in the Clearskies plugin for Claude runs the same workflow on demand: a prioritized risk view across your pipeline, the signal that flagged each deal, and the rep who owns it — all from your Context Graph.

​
Trigger phrases

  • “Show me pipeline risk”
  • “What deals are at risk this quarter”
  • “Where am I most exposed in the pipeline”

​
Customize it for your team

  • Match the risk signals to your team’s playbook
  • Adjust the prioritization (by deal value, close proximity, or stage)
  • Layer in coaching context — which reps need help with which deal types
Want this packaged for your team? Reach out and we’ll help you customize the skill to your workflow.
  • Deal Status — the on-demand read on a single deal when you need to answer “what changed”
  • Pipeline Review — the week-over-week movement read on a cadence, complementary to the always-on alerts here

​
The Plugin

Ready to make this part of your team’s workflow? We’ll set up the Clearskies plugin with you — Clearskies workflows, Claude skills, all tailored to how your team works. Book 15 minutes →

​
Next steps

  1. Sign in to Clearskies
  2. Connect your data sources
  3. Get your Clearskies MCP server and try with Claude or ChatGPT