Walk into the forecast call with a number you can defend.
Sorts each deal into commit, best-case, or pipeline by evidence, flags one risk, and scripts what to press each rep on.
Inside: The evidence-based sort, one risk per deal, and the press question.
Install it in one line, or paste it in.
~/.claude/skills/ and runs automatically when it is relevant.Connect your context. Set it to your motion.
it reads all of the above automatically, plus category history so it can see what moved.
confirms the last real touch and who is actually engaged.
adds trial usage momentum on deals that ride on a pilot.
This was built for a B2B SaaS org running a staged pipeline and a weekly forecast call. Set these to your stack:
Run any qualification method you like. The skill checks "does the evidence support the category this deal claims," so point it at your fields, not anyone else's.
| Set this | What it is | Default / Example |
|---|---|---|
| STAGE field | the opportunity stage | Opportunity.StageName |
| FORECAST field | commit / best-case / pipeline category | ForecastCategoryName |
| AMOUNT field | the number that rolls up | Opportunity.AmountARR |
| CLOSE field | the committed close date | Opportunity.CloseDate |
| ACTIVITY sources | where "last touch" lives | AccountOpportunityTaskEvent dates |
| PERIOD | the forecast window | this monththis quarter |
| THIN_DAYS | no-activity days that make a commit soft | 10 (re-tune to your cycle) |
Everything the skill does, in full.
Takes your open deals and builds the forecast call for you, deal by deal. It sorts each one into commit, best-case, or pipeline based on what the evidence supports, not what the category says. It flags the single biggest risk on each deal. Then it hands you the exact question to ask the rep so the call moves the number instead of narrating it.
- 1Category sort (three buckets, evidence-based)
Every deal lands in COMMIT, BEST_CASE, or PIPELINE based on what the evidence supports, then that is compared to the category the rep set.
- COMMIT: close date in the period, recent two-way activity, more than one contact engaged, next step set. A deal missing any of these is a soft commit and gets called out.
- BEST_CASE: real but not locked. A credible path to close in the period with at least one open risk.
- PIPELINE: possible, not this period. Named as upside, not counted on. The gap between the evidence bucket and the rep's category is the whole point. A COMMIT the evidence rates BEST_CASE is where the number breaks.
- 2One risk per deal
Every deal gets exactly one flagged risk, the one most likely to sink it: single-thread, no next step, close date already slipped once, champion quiet, thin activity, or legal/procurement not started. One risk, named, so the call is about that risk and nothing else.
- 3The question to press
Each deal gets one scripted question for the rep, aimed at the flagged risk. Not "how's it going." Something like "who besides your champion has said yes, and when did they say it?" The question is designed to surface the truth, not confirm the hope.
- 4Soft-commit watch
Flag any COMMIT where the evidence is thin: no activity in THIN_DAYS, a single thread, a close date that already moved, or a next step that is blank. These are the deals that quietly miss. Surface them before the call, not after the quarter.
- No deal called a soft commit without naming the specific thing that is missing.
- Every deal carries exactly one scripted question, tied to its one flagged risk.
- Close-date slips are shown with the prior date, never asserted.
- The number is presented three ways: rep-committed, evidence-supported, and the gap between them.
FORECAST CALL PREP · 6 deals in the period Account Category Evidence Amount Risk Press the rep on Acme Corp Commit Commit $45K Clean Confirm signature date Vertex Commit Best-case $67K Single-thread Who else has said yes? Blend Labs Best-case Best-case $28K No next step What is the next date? Northwind Commit Pipeline $112K Close slipped 2x Why is this month real? Number three ways: Rep-committed: $224K Evidence-supported: $73K Gap to defend: $151K (Vertex, Northwind) Walk in with: 1. Vertex. Get a second yes on record or move it to best-case. 2. Northwind. Third slip. Make the rep prove the date or push it out.
THIN_DAYS (10) and the "more than one contact engaged" cutoff are defaults, not laws. They suited a mid-market SaaS cycle with a weekly call. If your deals run longer or your call is monthly, raise them. The buckets do not change. The thresholds are yours.
Where an operator takes this next.
The prep works from a pasted export the morning of the call. Here is where it stops being a scramble.
The call stops being a status update and starts being where the number gets defended.
A scheduled Claude task pulls the pipeline export from Salesforce every Monday morning and has this prep waiting before the call starts.
Connect Slack so a deal that slips from Commit to Best-case mid-week posts to the deal channel the moment it happens, not at the next call.
Log the rep's answer back into the CRM against the flagged risk, so next quarter you can see which risks actually predicted a miss.
One skill is the on-ramp.
A single skill does one job. Chained into a playbook, or run as a full build, it becomes a system. Here is where this one plugs in.