Passed 16 of 22 test cases with the skill, and 12 of 22 without it.
Without skill55%
With skill73%
Checks it passed that the model alone missed
submit brand ownership documentation to the acquirer for active disputes
issue a sales allowance certificate to adjust tax output for the cross-period loss
state that a cross-period sales allowance certificate is not required when voiding within the same bimonthly tax period
and 1 more in the full report
4 fixed12 passed either way6 still failing0 broke
Pass-rate lift
+18 pp
Tokens
+19%
Agent turns
0%
Response time
+9%
Model gemini-3.6-flashMeasured 8/28/2026v116 of 22 cases passed with skill
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 4 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 4
Prompt
A customer places an order on March 10th with a Taiwan B2C e-commerce shop using a local credit card. On March 12th, before the item ships, the customer cancels the order via customer support. The invoice was issued on March 10th. What tax document procedure should be performed to reverse this transaction in the electronic invoice system?
What was checked, without and with the skill
Checked
Without skill
With skill
recommend performing a direct e-invoice void operation for the same-period transaction
Passed
Passed
state that a cross-period sales allowance certificate is not required when voiding within the same bimonthly tax period
Missed
Passed
case-02 · gemini-3.6-flash · 8/28/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md47 linessingle file
Payment Dispute Handling
STATUS: SKELETON — body pending.
When to use this skill
A customer has filed a credit card chargeback
Issuing a refund that crosses the bimonthly invoice boundary
Building an internal dispute-handling SOP
Preparing evidence packets for acquirer review
Reconciling dispute outcomes against invoice state