Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Issue, cancel, and fetch Hungarian invoices via the szamlazz.hu Agent API. Handles VAT calculation, NAV taxpayer lookup, partner caching, and PDF generation. Use when the user mentions számla, számlázás, invoice, sztornó, díjbekérő, proforma, or wants to bill a customer.
.claude/skills/davepoon-szamlazz-invoicing/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 31% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -19% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -28% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -3% | 0% |
| case-18 | ✗→✓ | ▲ Improved | 109% | 0% |
Issue, cancel (storno), download, and manage Hungarian invoices directly from Claude Code via the szamlazz.hu Agent API.
Built by SocialPro — a Hungarian AI automation and digital marketing agency specializing in custom AI integrations for SMEs.
pip install pyyaml)bash/plugin marketplace add socialproKGCMG/socialpro-plugins /plugin install szamlazz@socialpro-plugins
seller.yaml/szamlazz állíts ki egy számlát Példa Kft.-nek 150 000 Ft-ról webfejlesztésről/szamlazz sztornózd a SOC-2026-0042 számlát/szamlazz díjbekérő Acme Ltd-nek 500 EUR-ról konzultációért/szamlazz töltsd le a SOC-2026-0042 PDF-jét/szamlazz ki ez a cég: 12345678-2-42| Feature | Details | |---|---| | Invoice types | Regular, proforma (díjbekérő), storno | | VAT rates | 27% / 18% / 5% / 0% / AAM + KATA support | | NAV lookup | Auto-fetches company name and address from tax number | | Partner cache | Customers remembered by tax ID in local partners.yaml | | Cross-platform | macOS, Linux, Windows | | Rounding | Decimal with ROUND_HALF_UP to 2 decimals | | Error handling | 7 most common szamlazz.hu errors translated with recovery steps | | Security | API key in OS credential store, never echoed to stdout |
| Code | Meaning | Fix | |---:|---|---| | 3 | Auth failed | Regenerate Agent key at szamlazz.hu | | 54, 55 | e-Számla cert | Retry with eszamla=false | | 57, 259-264 | Calculation mismatch | Recalculate with Decimal rounding | | 136 | Unpaid balance | Pay szamlazz.hu subscription |
penznem and the MNB exchange rate is used automatically| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-22 | fail→fail | 26,142 | 26,485 | +1% | 1 | 1 | 0% | 6,181 | 7,347 | +19% | 0 | 0 | — |
case-01 | fail→fail | 30,248 | 13,154 | -57% | 1 | 1 | 0% | 6,184 | 1,369 | -78% | 0 | 0 | — |
case-02 | fail→fail | 7,254 | 5,314 | -27% | 1 | 1 | 0% | 1,235 | 1,432 | +16% | 0 | 0 | — |
case-03 | fail→pass | 7,915 | 4,858 | -39% | 1 | 1 | 0% | 1,423 | 1,867 | +31% | 0 | 0 | — |
case-16 | fail→pass | 10,405 | 37,532 | +261% | 1 | 1 | 0% | 1,662 | 1,346 | -19% | 0 | 0 | — |
case-04 | pass→pass | 9,121 | 3,386 | -63% | 1 | 1 | 0% | 1,670 | 1,686 | +1% | 0 | 0 | — |
case-05 | fail→fail | 5,832 | 3,137 | -46% | 1 | 1 | 0% | 911 | 1,454 | +60% | 0 | 0 | — |
case-06 | pass→pass | 5,999 | 5,594 | -7% | 1 | 1 | 0% | 964 | 2,078 | +116% | 0 | 0 | — |
case-07 | pass→pass | 3,024 | 1,325 | -56% | 1 | 1 | 0% | 488 | 1,302 | +167% | 0 | 0 | — |
case-08 | pass→pass | 7,885 | 4,502 | -43% | 1 | 1 | 0% | 1,162 | 1,908 | +64% | 0 | 0 | — |
case-09 | fail→pass | 11,823 | 2,082 | -82% | 1 | 1 | 0% | 1,975 | 1,431 | -28% | 0 | 0 | — |
case-10 | pass→pass | 12,675 | 1,678 | -87% | 1 | 1 | 0% | 2,083 | 1,398 | -33% | 0 | 0 | — |
case-11 | fail→fail | 7,685 | 1,639 | -79% | 1 | 1 | 0% | 1,342 | 1,395 | +4% | 0 | 0 | — |
case-12 | pass→pass | 10,082 | 3,613 | -64% | 1 | 1 | 0% | 1,799 | 1,858 | +3% | 0 | 0 | — |
case-13 | fail→pass | 9,157 | 2,054 | -78% | 1 | 1 | 0% | 1,516 | 1,471 | -3% | 0 | 0 | — |
case-14 | pass→pass | 7,824 | 3,098 | -60% | 1 | 1 | 0% | 1,398 | 1,681 | +20% | 0 | 0 | — |
case-15 | pass→pass | 11,922 | 2,969 | -75% | 1 | 1 | 0% | 2,540 | 1,599 | -37% | 0 | 0 | — |
case-21 | pass→fail | 9,040 | 5,515 | -39% | 1 | 1 | 0% | 1,561 | 2,039 | +31% | 0 | 0 | — |
case-17 | pass→pass | 8,329 | 2,076 | -75% | 1 | 1 | 0% | 1,437 | 1,456 | +1% | 0 | 0 | — |
case-18 | fail→pass | 4,586 | 1,828 | -60% | 1 | 1 | 0% | 698 | 1,461 | +109% | 0 | 0 | — |
case-19 | pass→pass | 6,697 | 3,415 | -49% | 1 | 1 | 0% | 1,060 | 1,662 | +57% | 0 | 0 | — |
case-20 | pass→pass | 7,611 | 6,700 | -12% | 1 | 1 | 0% | 1,271 | 2,301 | +81% | 0 | 0 | — |
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, and 20 counted toward the lift figure. The other 2 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +18 percentage points is the difference between those two pass rates over the 20 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
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.
Other measured skills in the registry, with their headline benchmark lift.