Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Fertig erstelltes Common-Law-Arbeitsprodukt auf Qualitaet prüfen: Jurisdiktion Quellenstand False Friends UK/US-Trennung Review-Bedarf. Prüfraster Jurisdiktion-Konsistenz Normen-Aktualitaet False-Friends-Scan UK-US-Trennung. Output Qualitaets-Prüfbericht Lueckenliste. Abgrenzung zu common-law-false-friends-scanner (Begriffe) und common-law-bilingual-contract-review (Vergleich).
.claude/skills/thomasmoreai-common-law-quality-gate/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 146% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 198% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 208% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 127% | 0% |
| case-20 | ✓→✗ | ▼ Worse | 46% | 0% |
Keine finale Freigabe ohne Jurisdiktion, verbindliche Sprache, UK/US-Trennung, Quellenstand und Review-Level.
Common-Law-Kompass arbeitet freundlich, präzise und verzeihend. Der Stil darf leicht sein, aber nie auf Kosten der juristischen Trennschärfe.
Kläre vor dem Qualitaets-Gate-Durchlauf:
UK
US
Kommentar
| Pruefpunkt | Standard | Fundstelle | Status | |---|---|---|---| | Governing law & jurisdiction | Eindeutige Rechtswahlklausel | Spiliada 1987] AC 460 | GRUEN/GELB/ROT | | Consideration (UK) | Nicht illusorisch, nicht past | Currie v Misa (1875) LR 10 Ex 153 | — | | Entire agreement clause | Vollstaendig, UCTA-konform | UCTA 1977 s 3 | — | | Limitation of liability | Reasonableness-Test bestanden | UCTA 1977 s 11 | — | | Indemnity-Kaskaden | Gegenseitig oder cap-gebunden | Caledonian North Sea 2002] 1 Lloyd's Rep 553 | — | | Consequential loss exclusion | Direct/indirect klar definiert | Hadley v Baxendale (1854) 9 Ex 341 | — | | Force majeure | Ereignis, Notice, Dauer, Exit | Seadrill Ghana 2018] EWHC 1640 | — | | Arbitration/dispute clause | Vollstaendig, seated, institutional | Fiona Trust 2007] UKHL 40 | — |
Adressat: Partnerebene / Senior Associate — Tonfall: strukturiert, risikofokussiert
COMMON-LAW QUALITY-GATE-MEMO
Datum: [DATUM]
Dokument: [VERTRAGSNAME / VERSION]
Governing Law: [RECHTSORDNUNG]
Reviewer: [NAME]
AMPEL-GESAMTBEWERTUNG: GRUEN / GELB / ROT
KRITISCHE BEFUNDE (ROT)
1. [Klausel / Abschnitt]: [Befund] — Risiko: [Beschreibung] — Empfehlung: [Konkrete Aenderung]
KLAERUNGSBEDARF (GELB)
1. [Klausel / Abschnitt]: [Befund] — Empfehlung: [...]
NICHT-BEANSTANDET (GRUEN)
- [Liste der geprueften und in Ordnung befundenen Punkte]
OFFENE ANNAHMEN
- [...]
LOCAL-COUNSEL-BEDARF
- [Ja/Nein; wenn ja: fuer welche Punkte]
NAECHSTE SCHRITTE
- [Schritt mit Frist und Verantwortlichem]| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 19,927 | 11,175 | -44% | 1 | 1 | 0% | 3,381 | 3,921 | +16% | 0 | 0 | — |
case-02 | fail→fail | 17,598 | 20,903 | +19% | 1 | 1 | 0% | 3,378 | 5,710 | +69% | 0 | 0 | — |
case-03 | fail→pass | 11,243 | 17,819 | +58% | 1 | 1 | 0% | 2,084 | 5,117 | +146% | 0 | 0 | — |
case-04 | pass→pass | 18,072 | 15,387 | -15% | 1 | 1 | 0% | 3,069 | 4,632 | +51% | 0 | 0 | — |
case-05 | pass→pass | 11,392 | 16,250 | +43% | 1 | 1 | 0% | 2,094 | 4,653 | +122% | 0 | 0 | — |
case-06 | pass→pass | 11,573 | 14,385 | +24% | 1 | 1 | 0% | 2,251 | 4,433 | +97% | 0 | 0 | — |
case-07 | pass→pass | 19,056 | 19,138 | +0% | 1 | 1 | 0% | 3,530 | 5,148 | +46% | 0 | 0 | — |
case-08 | pass→pass | 14,735 | 17,708 | +20% | 1 | 1 | 0% | 2,782 | 5,022 | +81% | 0 | 0 | — |
case-09 | fail→pass | 8,923 | 15,071 | +69% | 1 | 1 | 0% | 1,397 | 4,158 | +198% | 0 | 0 | — |
case-10 | fail→fail | 14,836 | 16,551 | +12% | 1 | 1 | 0% | 2,601 | 4,473 | +72% | 0 | 0 | — |
case-11 | pass→pass | 11,513 | 15,373 | +34% | 1 | 1 | 0% | 2,237 | 4,453 | +99% | 0 | 0 | — |
case-12 | fail→pass | 11,766 | 15,828 | +35% | 1 | 1 | 0% | 1,466 | 4,519 | +208% | 0 | 0 | — |
case-13 | pass→pass | 13,426 | 16,672 | +24% | 1 | 1 | 0% | 2,720 | 4,773 | +75% | 0 | 0 | — |
case-14 | pass→pass | 15,756 | 20,770 | +32% | 1 | 1 | 0% | 3,176 | 5,662 | +78% | 0 | 0 | — |
case-15 | pass→pass | 15,530 | 22,457 | +45% | 1 | 1 | 0% | 2,917 | 5,493 | +88% | 0 | 0 | — |
case-16 | fail→pass | 12,262 | 17,329 | +41% | 1 | 1 | 0% | 2,042 | 4,639 | +127% | 0 | 0 | — |
case-17 | pass→pass | 16,140 | 11,528 | -29% | 1 | 1 | 0% | 2,903 | 3,703 | +28% | 0 | 0 | — |
case-18 | pass→pass | 18,399 | 24,386 | +33% | 1 | 1 | 0% | 3,191 | 5,867 | +84% | 0 | 0 | — |
case-19 | pass→pass | 22,555 | 29,485 | +31% | 1 | 1 | 0% | 4,206 | 7,288 | +73% | 0 | 0 | — |
case-20 | pass→fail | 20,430 | 19,464 | -5% | 1 | 1 | 0% | 3,863 | 5,657 | +46% | 0 | 0 | — |
case-21 | pass→pass | 24,224 | 16,874 | -30% | 1 | 1 | 0% | 3,310 | 4,807 | +45% | 0 | 0 | — |
case-22 | pass→pass | 13,145 | 20,170 | +53% | 1 | 1 | 0% | 2,306 | 4,422 | +92% | 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. The headline lift of +14 percentage points is the difference between those two pass rates over the 22 comparable cases. 2 cases got worse with the skill loaded, and they are 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.