SSkillEngineering
Honest Metrics Audit (Claude Code)
claude codeproducttrustreviewfree
Sample outputCaptured sample. Not a live run.
CRITICAL Active users
Shows: users.length
Counts: rows in the database
Claims: distinct active users
Visible: homepage
Fix: users.filter(distinct).length or relabel to "database rows"
HIGH Verified sellers
Shows: items.filter(i => i.seller.verified).length
Counts: items
Claims: verified sellers
Visible: sellers page
Fix: relabel to "verified items" or change expression to sellers.filter(verified).length
MEDIUM Downloads
Shows: downloads.toLocaleString()
Counts: downloads
Claims: downloads
Visible: product page
Fix: add unit label for clarity, e.g. "downloads (total)"
Total numbers traced: 10
Total matches: 7
Total mismatches: 3
Generated from this listing's real file at upload. The full output is in your download.
⛨Free, and yours to run in your own tools. No account needed.
TypeSkillCategoryEngineeringRuns onClaude CodeLicenseStandard license
⛨Human-reviewedEvery listing checked before it goes live
⛨Source-stampedCarries a visible line saying where it came from
⛨Real reviewsOnly verified buyers can rate it
A Claude Code skill that traces each displayed figure back to the query behind it. Nobody sets out to publish a false number: someone counts items, someone else labels it people, and both lines are individually reasonable. Together they tell every visitor there are fifteen verified sellers when there is one, and no test fails because no test knows what the label says. Catches counts of the wrong entity, missing distinct, populations that include your own test accounts, wrong time windows, flattering denominators, and hardcoded traction that arrived as design placeholder text. Also checks what renders when the real value is zero or the query fails, because a stat that falls back to a plausible default is asserting something nothing supports.
- ✓1 .md (SKILL.md) file (6 KB), instant download
- ✓Runs on Claude Code
- ✓Version 1.0.0, all v1.x updates included free