AI Skills Marketplace / CRM & RevOps

CRM Pipeline Hygiene Auditor

The CRM Pipeline Hygiene Auditor is a free Claude Skill that flags stale deals against per-stage-velocity thresholds, missing required fields, fuzzy-match duplicate deals, and stage-skip anomalies, output as a single prioritized cleanup list.

↗ View Source ⬇ Download Full Repo (.zip)
MIT
License
4 hygiene checks
Decision branches
per-stage velocity thresholds
Escalation rules
Yes
Multi-tool support
Aug 14, 2026
Last updated
15/15 passing
Tests
README.md View raw ↗

What this is (and isn’t)

An audit function you run against exported deal data, not a hosted service or a CRM integration itself. It runs entirely in your own environment.

Files

FilePurpose
auditor.jsauditPipeline(), plus exported levenshtein()/similarity() helpers — zero dependencies
test.js15-test suite covering staleness, missing fields, dupe detection, and stage-skip logic
SKILL.mdFull skill definition, loadable in Claude Code/Desktop

Why this is a real gap

A fixed "flag deals stale after 30 days" rule treats a fast-moving Prospecting stage the same as a naturally slow Negotiation stage. This skill compares each deal’s actual days-in-stage against your own historical median for that specific stage.

FAQ

Is this skill free to use?

Yes. Free to download and run yourself. MV3 charges $175/hr only if you want it wired into your real CRM so the audit runs automatically.

Which CRMs does this work with?

Any CRM you can export deal data from, including Salesforce, HubSpot, Pipedrive, and Zoho, all of which have real, shipped MCP servers as of this research.

How does the staleness threshold work?

It compares each deal's days-in-stage against your own historical median for that stage times a configurable multiplier, not a fixed day count applied to every stage equally.

Does my CRM data get sent to MV3?

No. The audit runs entirely in your own environment.

What counts as a possible duplicate?

Deal pairs with combined name and company similarity above a configurable threshold (fuzzy string matching, not exact match), so near-identical entries are caught even with small typos or formatting differences.

JR
Jordan Reeves
ABM & Outbound Pipeline, MV3 Marketing