Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

GA4 for Taiwan E-Commerce · asgard-ai-platform | DecimalAI

asgard-ai-platform/tw-ecom-analytics-ga4

NewSkillScoreMaintained

Implement GA4 for Taiwan e-commerce — Enhanced Ecommerce events (view_item, add_to_cart, begin_checkout, purchase), TW-specific parameter conventions (含稅 revenue, NT$ currency, 統編 as user property), Looker Studio reporting, and Big Query export. Use when instrumenting a TW store with GA4 or auditing existing GA4 setup. Do NOT use for generic analytics (use `ecom-analytics`). STATUS: SKELETON — body pending.

by asgard-ai-platform

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Data Analyst pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/asgard-ai-platform-tw-ecom-analytics-ga4/SKILL.md

decimalai skills pull asgard-ai-platform-tw-ecom-analytics-ga4 --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/asgard-ai-platform-tw-ecom-analytics-ga4/SKILL.md https://app.decimal.ai/s/asgard-ai-platform-tw-ecom-analytics-ga4/SKILL.md

Raw file: /s/asgard-ai-platform-tw-ecom-analytics-ga4/SKILL.md ↗ · machine-readable summary at /s/asgard-ai-platform-tw-ecom-analytics-ga4.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 31h ago

Yes — +5% lift+17% tokens0% turns91% of whole cases pass with skill (20/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-04	✗→✓	▲ Improved	16%	0%
case-01	✓→✓	＝ Same ✓	66%	0%
case-02	✓→✓	＝ Same ✓	1%	0%
case-03	✓→✓	＝ Same ✓	32%	0%
case-05	✓→✓	＝ Same ✓	44%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/3/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

asgard-ai-platform / skills

tw-ecom-analytics-ga4/SKILL.md

View on GitHub ↗

License MITImported 8/3/2026Version v1

Skill contents

SKILL.md

GA4 for Taiwan E-Commerce

> STATUS: SKELETON — body pending.

When to use this skill

Instrumenting a TW store with GA4 Enhanced Ecommerce
Auditing an existing GA4 setup for TW conventions
Designing Looker Studio reports for TW KPI
Setting up BigQuery export
Attributing across LINE OA / marketplace channels

Do NOT use when

Generic analytics → ecom-analytics
Benchmarks only → tw-ecom-analytics-benchmarks

Core concepts

TODO: GA4 event model, 含稅 revenue handling, currency = TWD, content_group usage for 檔期.

Decision tree

TODO: event → parameter mapping for TW conventions.

Implementation guidance

TODO: dataLayer template, tag setup, consent mode, BigQuery export.

Gotchas

TODO: 5-6 pitfalls (含稅 double-count, cross-domain marketplace attribution, LINE IAB tracking block, consent-mode revenue undercount, parameter cardinality limits).

IRON LAW

TODO.

Output Format

TODO.

ecom-analytics
tw-ecom-analytics-benchmarks

_Last verified: 2026-04_

Want to see if asgard-ai-platform/tw-ecom-analytics-ga4 would activate on your phrasing? Try it before you install — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/28/2026, 06:42 UTC · 22 cases

Lift

+5%

91% vs 86%

Agent turns

1.0 vs 1.0

Tokens

+17%

2,720.77 vs 2,326.64

fail → pass · 1pass → pass · 19fail → fail · 2

Per-case results (22)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-01 I am setting up GA4 purchase dataLayer events for an online clothing store based in Taipei selling exclusively to local customers. The developer wants to use the default currency code USD since their payment gateway API returns amounts in cents. Specify the correct ISO 4217 currency code to use for all GA4 ecommerce events in Taiwan.	pass→pass	6,555	10,594	+62%	1	1	0%	1,225	2,036	+66%	0	0	—
case-02 In Taiwan, product prices shown on e-commerce storefronts are legally required to include 5% VAT. When sending the standard GA4 purchase event dataLayer payload for a 1,050 TWD order (1,000 product + 50 tax), should the value parameter represent the tax-inclusive total (1,050 TWD) or net total (1,000 TWD), and how should the 5% tax be passed?	pass→pass	16,017	14,475	-10%	1	1	0%	2,922	2,938	+1%	0	0	—
case-03 Our Taiwan store gets 40% of its traffic from LINE Official Account broadcast links. In GA4 reports, these visitors are frequently recorded as direct traffic or lost sessions when they switch from LINE's internal browser to Chrome or Safari to complete payment. What URL parameter or strategy should be appended to LINE links to preserve GA4 campaign attribution and session continuity?	pass→pass	17,170	19,439	+13%	1	1	0%	2,437	3,224	+32%	0	0	—
case-04 During major Taiwan shopping festivals like Double 11 and Mother's Day, our marketing team wants to aggregate GA4 ecommerce performance across different promotional landing pages. How should campaign promotional themes be structured using GA4 native parameters?	fail→pass	19,642	22,291	+13%	1	1	0%	3,487	4,050	+16%	0	0	—
case-05 Write a BigQuery SQL query against GA4 exported data tables to calculate daily total sales for a Taiwan e-commerce store. The query must correctly extract transaction revenue from the nested parameters array for purchase events.	pass→pass	9,711	13,495	+39%	1	1	0%	1,942	2,787	+44%	0	0	—
case-06 To comply with Google's updated ad policies for GA4 remarketing and tracking in Taiwan, our site needs Google Consent Mode v2 configured via GTM. Which default consent parameters must be declared before any GA4 tags fire?	pass→pass	8,906	11,052	+24%	1	1	0%	1,634	2,149	+32%	0	0	—
case-07 Provide the dataLayer push JavaScript code for a GA4 product addition event when a customer in Taiwan adds a 2,500 TWD mechanical keyboard (variant ID KB-99, category Peripherals) to their cart.	fail→fail	10,530	7,643	-27%	1	1	0%	1,636	1,773	+8%	0	0	—
case-08 When a user in Taiwan clicks the checkout button on an online store with a total cart value of 3,200 TWD, what is the standard GA4 event structure for initiating checkout?	pass→pass	7,790	9,389	+21%	1	1	0%	1,504	2,017	+34%	0	0	—
case-09 We are building a Looker Studio executive dashboard connected to GA4 for a Taiwan e-commerce brand. The total revenue field currently displays with a standard $ symbol. How should the field formatting be configured in Looker Studio for Taiwan currency display?	pass→pass	12,298	9,220	-25%	1	1	0%	2,117	1,796	-15%	0	0	—
case-10 Our Taiwan retail site displays product cards on category pages. Which GA4 Enhanced Ecommerce event should be fired when product items are rendered in the viewport, and what item list parameter tracks the collection name?	pass→pass	8,993	7,817	-13%	1	1	0%	1,684	1,735	+3%	0	0	—
case-11 Our analytics team wants to send the customer's full Taiwan postal delivery address (including house number and floor) as a custom event parameter on every GA4 purchase event. What potential GA4 operational risk does this introduce, and how should address details be scoped?	pass→pass	15,113	14,311	-5%	1	1	0%	2,224	2,740	+23%	0	0	—
case-12 A Taiwan brand runs ads sending traffic to both their self-hosted DTC site and their official Shopee store. Marketers want to track seamless cross-domain GA4 user journeys between DTC and Shopee. What technical limitation prevents native GA4 cross-domain session linking with Shopee?	fail→fail	10,866	12,049	+11%	1	1	0%	1,800	1,992	+11%	0	0	—
case-13 Write a SQL snippet for GA4 BigQuery export data to extract item-level details (item_id, item_name, price, discount) from the repeated record field inside purchase events.	pass→pass	8,057	9,297	+15%	1	1	0%	1,481	1,752	+18%	0	0	—
case-14 When a shopper on a Taiwan store clicks a recommended product card on the homepage, which GA4 event should be dispatched to track item interaction before page navigation?	pass→pass	8,233	6,357	-23%	1	1	0%	1,364	1,185	-13%	0	0	—
case-15 A customer in Taiwan returns an order worth 1,500 TWD due to sizing issues. What is the standard GA4 event payload structure to process a full refund in analytics?	pass→pass	11,748	14,631	+25%	1	1	0%	2,203	2,456	+11%	0	0	—
case-16 In Taiwan e-commerce, customers frequently choose between Credit Card, ATM transfer, Line Pay, and Convenience Store COD. How should payment method selection be recorded in GA4 during checkout?	pass→pass	16,256	13,423	-17%	1	1	0%	2,963	2,640	-11%	0	0	—
case-17 Taiwan online stores commonly offer convenience store pickup versus home delivery. How should shipping option selection be tracked in GA4?	pass→pass	16,285	18,663	+15%	1	1	0%	2,883	3,058	+6%	0	0	—
case-18 To combine offline CRM loyalty data with online GA4 behavior for Taiwan shoppers, how should authenticated customer IDs be sent to GA4 without violating Google PII policies?	pass→pass	17,765	25,373	+43%	1	1	0%	2,840	3,913	+38%	0	0	—
case-19 When building a Looker Studio dashboard for a Taiwan DTC shop, how should E-commerce Conversion Rate be calculated from GA4 metrics?	pass→pass	16,039	16,320	+2%	1	1	0%	2,590	2,820	+9%	0	0	—
case-20 We are launching a US-focused Shopify store targeting customers in California selling in USD. We need advice on general Shopify store setup and US sales tax nexus configuration for California state taxes.	pass→pass	16,706	19,569	+17%	1	1	0%	2,771	3,358	+21%	0	0	—
case-21 What are the current industry benchmark average conversion rates, customer acquisition costs, and average order values for online apparel brands versus consumer electronics brands in Taiwan?	pass→pass	18,338	18,483	+1%	1	1	0%	2,785	3,494	+25%	0	0	—
case-22 How do we implement Meta Conversions API server-side event tracking using Node.js for a Taiwan WooCommerce store?	pass→pass	31,339	34,870	+11%	1	1	0%	4,694	5,944	+27%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +5 percentage points is the difference between those two pass rates over the 22 comparable cases.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 86.4%
Pass rate with the skill: 90.9%
Cases attempted: 22
Cases counted toward lift: 22
Cases that regressed: 0
Skill version measured: v1
Run completed: 8/28/2026, 06:42 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-04

Prompt

During major Taiwan shopping festivals like Double 11 and Mother's Day, our marketing team wants to aggregate GA4 ecommerce performance across different promotional landing pages. How should campaign promotional themes be structured using GA4 native parameters?

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signalsno signals yet

BenchmarkRun a verified benchmark with positive lift

Live evalNeeds ≥20 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

No signals yet — run a benchmark or collect evaluated traces to earn a score.

Verified on gemini-3.6-flash · re-tested yesterday

Other measured skills in the registry, with their headline benchmark lift.