Solve "Applicable Fee IDs" questions in the dabstep payment dataset. Use this skill whenever a question asks which fee IDs apply to a merchant, transaction, or combination of payment attributes (account_type, aci, card_scheme, capture_delay, intracountry, MCC, etc.). Covers single-day queries ("what fees apply on day 200 for MerchantX?"), single-month queries ("what fees applied in October 2023?"), and full-year queries ("what fees applied in 2023?").
Standardmäßig ist der Prompt ausgewählt, der zuerst die Quelle prüft. Sie können zu einem direkten Befehl wechseln oder eine lokale Kopie herunterladen.
Quelldateien prüfen
Lesen Sie SKILL.md und alle von SkillsMP angezeigten Begleitdateien, bevor Sie sich für eine Installation entscheiden.
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Ein direkter Befehl überspringt den Prüf-Prompt. Prüfen Sie die Quelle, bevor Sie ihn ausführen.
Solve "Applicable Fee IDs" questions in the dabstep payment dataset. Use this skill whenever a question asks which fee IDs apply to a merchant, transaction, or combination of payment attributes (account_type, aci, card_scheme, capture_delay, intracountry, MCC, etc.). Covers single-day queries ("what fees apply on day 200 for MerchantX?"), single-month queries ("what fees applied in October 2023?"), and full-year queries ("what fees applied in 2023?").
Applicable Fee IDs — Dabstep Dataset
Query Types
Type A — Single Day: "For the Nth of year 2023, what are the Fee IDs applicable to [Merchant]?"
Monthly stats = full calendar month containing day N
Transaction filter = only transactions on day N
Type B — Single Month: "What were the applicable Fee IDs for [Merchant] in [Month] 2023?"
Monthly stats = that calendar month
Transaction filter = all transactions in that month
Type C — Full Year: "What are the applicable fee IDs for [Merchant] in 2023?"
Iterate over each calendar month; union all applicable IDs across all months
Data Files
File
Purpose
fees.json
1000 fee rules; each rule is a dict with matching conditions + fixed_amount + rate
{"ID":42,"card_scheme":"GlobalCard",// always a specific scheme (never null)"account_type":["F","S"],// list or [] (empty = all)"capture_delay":"<3",// string or null"monthly_fraud_level":">8.3%",// string or null"monthly_volume":"100k-1m",// string or null"merchant_category_code":[5812],// list or [] (empty = all)"is_credit":true,// bool or null"aci":["A","C"],// list or [] (empty = all)"intracountry":true// bool or null (also stored as 1.0/0.0)}
Null / empty-list semantics (critical):null or [] means the rule applies to all values of that field.
Core Matching Code
import json, pandas as pd, datetime
fees = json.load(open("fees.json"))
merchants = json.load(open("merchant_data.json"))
payments = pd.read_csv("payments.csv")
# --- Merchant lookup ---
m = next(x for x in merchants if x["merchant"] == "MerchantName")
account_type = m["account_type"]
mcc = m["merchant_category_code"]
capture_delay = map_capture_delay(m["capture_delay"])
# --- Category helpers ---defmap_capture_delay(raw):
if raw in ("immediate", "manual"): return raw
days = int(raw)
if days < 3: return"<3"if days <= 5: return"3-5"return">5"defvolume_category(eur):
if eur < 100_000: return"<100k"if eur < 1_000_000: return"100k-1m"if eur < 5_000_000: return"1m-5m"return">5m"deffraud_category(rate_pct):
if rate_pct < 7.2: return"<7.2%"if rate_pct < 7.7: return"7.2%-7.7%"if rate_pct <= 8.3: return"7.7%-8.3%"return">8.3%"# --- Fee matching ---deffee_matches(fee, card_scheme, is_credit, aci, intracountry,
vol_cat, fraud_cat, account_type, mcc, capture_delay):
if fee["card_scheme"] != card_scheme: returnFalseif fee["account_type"] and account_type notin fee["account_type"]: returnFalseif fee["capture_delay"] isnotNoneand fee["capture_delay"] != capture_delay: returnFalseif fee["monthly_volume"] isnotNoneand fee["monthly_volume"] != vol_cat: returnFalseif fee["monthly_fraud_level"] isnotNoneand fee["monthly_fraud_level"] != fraud_cat: returnFalseif fee["merchant_category_code"] and mcc notin fee["merchant_category_code"]: returnFalseif fee["is_credit"] isnotNoneand fee["is_credit"] != is_credit: returnFalseif fee["aci"] and aci notin fee["aci"]: returnFalseif fee["intracountry"] isnotNoneandbool(fee["intracountry"]) != intracountry: returnFalsereturnTrue
Capture Delay Mapping
Merchant value
Fee rule category
"immediate"
"immediate"
"manual"
"manual"
"1" or "2" (days < 3)
"<3"
"3", "4", "5"
"3-5"
"7" or any value > 5
">5"
Monthly Stats Computation
Monthly stats (volume + fraud rate) are always computed over a full calendar month. Use datetime to get exact day-of-year ranges:
# Determine containing month, compute stats for that month
day_date = datetime.date(2023, 1, 1) + datetime.timedelta(days=N-1)
month = day_date.month
vol_cat, fraud_cat, month_txns = compute_monthly_stats(payments, merchant_name, 2023, month)
# Filter transactions for the specific day only
day_txns = payments[(payments["merchant"]==merchant_name) &
(payments["year"]==2023) & (payments["day_of_year"]==N)]
applicable = set()
if vol_cat andnot day_txns.empty:
day_txns = day_txns.copy()
day_txns["intracountry"] = day_txns["issuing_country"] == day_txns["acquirer_country"]
for _, row in day_txns.iterrows():
for fee in fees:
if fee_matches(fee, row["card_scheme"], row["is_credit"], row["aci"],
bool(row["intracountry"]), vol_cat, fraud_cat,
account_type, mcc, capture_delay):
applicable.add(fee["ID"])
Type B: Single Month
vol_cat, fraud_cat, txns = compute_monthly_stats(payments, merchant_name, 2023, month_number)
applicable = set()
if vol_cat andnot txns.empty:
txns = txns.copy()
txns["intracountry"] = txns["issuing_country"] == txns["acquirer_country"]
for _, row in txns.iterrows():
for fee in fees:
if fee_matches(fee, row["card_scheme"], row["is_credit"], row["aci"],
bool(row["intracountry"]), vol_cat, fraud_cat,
account_type, mcc, capture_delay):
applicable.add(fee["ID"])
Type C: Full Year
applicable = set()
for month inrange(1, 13):
vol_cat, fraud_cat, txns = compute_monthly_stats(payments, merchant_name, 2023, month)
if vol_cat isNoneor txns.empty: continue
txns = txns.copy()
txns["intracountry"] = txns["issuing_country"] == txns["acquirer_country"]
for _, row in txns.iterrows():
for fee in fees:
if fee_matches(fee, row["card_scheme"], row["is_credit"], row["aci"],
bool(row["intracountry"]), vol_cat, fraud_cat,
account_type, mcc, capture_delay):
applicable.add(fee["ID"])
Output Format
Return fee IDs as a sorted, comma-separated list:
29, 36, 51, 64, 65, 89, 107, ...
If no fees match, return an empty string "".
Common Pitfalls
[] means "all values apply" — never treat empty account_type, aci, or merchant_category_code lists as "no match".
card_scheme is never null — always an exact match (GlobalCard, NexPay, TransactPlus, SwiftCharge).
Volume/fraud matching is categorical — compute the merchant's category string, then compare with the fee rule's string using equality. Do NOT parse ranges numerically.
capture_delay mapping — convert merchant's raw value (e.g. "1") to the fee-rule category string ("<3").
intracountry type — fees.json stores as 1.0/0.0; always wrap with bool() before comparing.
Monthly stats scope — always compute volume/fraud over the full calendar month, even for single-day queries.
