Claude Skill

Forecast Sanity Checker

Flags suspicious deals in your forecast — stuck too long, suspect month-end growth, missing signals.

Download skill (.zip)Or download whole pack

What it does

Before submitting your forecast, this skill reviews your deals for the patterns that usually mean "this won't close when you think it will." Surfaces deals stuck in stage too long, deals that suspiciously grew near month-end, deals missing key buying signals. Helps you submit a forecast you can defend.

When to use

✓Before submitting your monthly or quarterly forecast
✓When your manager pushes back: "are you sure about Acme?"
✓You're new to a territory and want a sanity-check before committing numbers

When not to use

✗Mid-quarter pipeline reviews — too aggressive then, holds you to higher standard than the conversation requires

Install

Download the .zip, then unzip into your Claude skills folder.

mkdir -p ~/.claude/skills
unzip ~/Downloads/forecast-sanity-checker.zip -d ~/.claude/skills/

# Restart Claude Code session.
# Skill is now available — Claude will use it when relevant.

SKILL.md

---
name: forecast-sanity-checker
description: Reviews sales pipeline forecasts for suspect patterns — stuck deals, month-end growth, missing signals.
---

# Forecast Sanity Checker

When the user submits deals for forecast review, flag the ones with suspect patterns.

## Required input

For each deal in the forecast:
- Company
- Stage
- Deal size
- Days in current stage
- Expected close date
- Last touchpoint (date + outcome)
- Confidence (commit / best case / pipeline)

## Suspect patterns to flag

### 1. Stage decay
Deal in current stage too long given the typical stage velocity. Surface deals that are 2x the typical stage duration.

### 2. Suspect month-end growth
Deal size grew between two reviews without a stated reason (new use case discovered, more seats, etc.). Often indicates pressure to hit quota.

### 3. Last-touch staleness
Last touchpoint is more than 14 days ago for a deal forecast to close in 30 days. Cold deals don't suddenly warm up at month-end.

### 4. Missing buying signals
Deal in late-stage (proposal, negotiation) but missing one of: economic buyer engaged, budget confirmed, timeline driver named, technical evaluation done. If 2+ are missing, deal is softer than it looks.

### 5. Single-threaded
Late-stage deal with only one stakeholder engaged from the prospect side. Risk of champion leaving / changing roles.

### 6. Pricing not yet stated
Late-stage deal where the price hasn't been quoted to the prospect yet. Often means rep is afraid of the price reaction — deal will likely die at quote time.

## Output

For each flagged deal:
- **Deal name + size**
- **Pattern flagged** (one of the above)
- **Specific concern** (one sentence)
- **What to do this week** (one action that would either advance the deal or disqualify it cleanly)

For deals that look clean, say so:
- **Deals that look solid:** [list, brief]

## Constraints

- Don't flag every deal. Only the genuinely suspect ones.
- If the user is over-flagged on a forecast cycle, calibrate — say "of N deals, X are flagged. That's roughly average / high / low for a healthy pipeline."
- Be direct. "This deal looks soft because [specific reason]" — not "you might want to look at this one."

Example prompts

Once installed, try these prompts in Claude:

Forecast review. [paste deal list with stages, sizes, expected close dates, last touchpoint]. Flag what looks soft.

Related prompts

Don't want to install a skill? These prompts in /prompts cover similar ground for one-shot use:

analysis prompts