CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

tracking-leak-detector

Detect gaps between expected and actual conversion tracking across ad platforms (Meta, Google, LinkedIn, TikTok, Reddit) and CRMs. Surfaces blind spots in pixel/CAPI/Enhanced Conversions that silently erode ROAS measurement.

Quality

75%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/tracking-leak-detector/SKILL.md
SKILL.md
Quality
Evals
Security

Tracking Leak Detector

Detects gaps between expected and actual conversion tracking across all ad platforms and CRMs. Surfaces blind spots that silently erode ROAS measurement.


When to Use This Skill

  • User asks "why are my conversions low?" or "is my tracking working?"
  • ROAS numbers seem unexpectedly low or inconsistent
  • A platform is spending money but showing 0 conversions
  • CRM shows more/fewer conversions than ad platforms report
  • Conversions suddenly dropped on a platform
  • User wants to audit their tracking setup before scaling spend
  • User mentions "tracking leaks" or "attribution gaps"

Agent Tools

This skill provides two agent tools:

detect_tracking_leaks

Analyzes the last N days of data to find:

Leak TypeWhat It Detects
zero_conversion_spendPlatforms spending money but reporting 0 conversions
crm_ad_mismatchCRM conversion count differs significantly from ad platform totals
conversion_dropSudden drop (>70%) in conversions vs prior period
utm_missingCRM contacts from paid sources without UTM campaign data
view_through_blind_spotPlatforms (Meta, LinkedIn, TikTok) with no impression tracking

Each leak includes a severity (critical/warning/info), description, estimated impact, and specific recommendation.

attribution_health_check

Comprehensive check of the full attribution pipeline:

CheckWhat It Validates
CRM ConnectionsHubSpot, Attio, Salesforce token status and data freshness
Ad Platform DataData freshness for all connected platforms
Conversion EventsWhether conversion events are being recorded
View-Through TrackingWhether impression/view events exist for VTC attribution
UTM CoveragePercentage of CRM contacts with campaign attribution

How Leaks Are Detected

1. Zero-Conversion Spend (Critical)

IF platform.spend > $100 AND platform.conversions == 0
THEN → Likely broken pixel or misconfigured conversion action

2. CRM-to-Ad Mismatch (Warning)

IF crm_conversions / ad_conversions > 1.5
THEN → Ads are under-reporting (UTM tracking gaps)

IF crm_conversions / ad_conversions < 0.5
THEN → Ads are over-counting (double-counting or CRM sync broken)

3. Conversion Drop (Critical)

IF current_period_conversions / prior_period_conversions < 0.3
THEN → 70%+ drop suggests broken tracking, not performance decline

4. UTM Missing (Warning)

IF contacts_from_paid_without_utm / total_contacts > 0.2
THEN → 20%+ of paid contacts have no campaign attribution

5. View-Through Blind Spot (Info)

IF platform IN (Meta, LinkedIn, TikTok, Pinterest)
   AND impression_events == 0
THEN → View-through conversions are not being captured

Supported CRMs

CRMRevenue SourceStatus
HubSpotDeal amounts from crm_events (deal_won)Active
AttioDeal/record values synced via APIActive
SalesforceOpportunity amountsActive

Supported Ad Platforms

PlatformConversion TrackingView-Through
Google AdsClick-throughN/A (handled by Google)
LinkedIn AdsClick-through + VTCSupported
Meta AdsClick-through + VTCSupported
Microsoft AdsClick-throughN/A
Reddit AdsClick-throughPlanned
X AdsClick-throughPlanned
TikTok AdsClick-through + VTCSupported
Amazon AdsClick-through + VTCSupported
StackAdaptClick-through + VTCSupported
The Trade DeskClick-through + VTCSupported

Example Usage

User: "My ROAS seems way too low on LinkedIn. Is something wrong with tracking?"

Agent should:

  1. Call detect_tracking_leaks with days=30
  2. Call attribution_health_check
  3. Report any leaks found, focusing on LinkedIn
  4. If view-through blind spot detected, explain that LinkedIn VTCs may be uncounted
  5. Provide specific remediation steps

User: "Run a tracking audit before I increase spend"

Agent should:

  1. Call attribution_health_check first for overall status
  2. Call detect_tracking_leaks for active issues
  3. Present a summary table of all findings
  4. Prioritize critical issues that should be fixed before scaling

Related Skills

  • conversion-tracking — Setup and verification of conversion pixels
  • 11-ad-platform-data — Pulling performance data from platforms
  • budget-optimizer — Budget allocation (depends on accurate attribution)

Last Updated: March 2026


About this skill

Part of the Synter free skills collection — open-source agent skills for advertising, PPC, and marketing automation.

Want this skill (and 40+ more) running on autopilot against your live ad accounts? Try Synter — AI Agent Media Buyers that connect to Google, Meta, LinkedIn, TikTok, Reddit, Amazon, and 7+ more platforms.

Repository

Synter-Media-AI/free-skills

Path

skills/tracking-leak-detector/SKILL.md

Last updated
First committed

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.