Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Anwalt oder Übersetzer sucht missverstaendliche deutsch-englische Rechtsbegriffe im Vertragstext oder Memo. Anwendungsfall Vertragsentwurf mit False-Friend-Risiko. Prüfraster Begriff-Scan Risikoeinstufung sichere Alternativen Jurisdiktion-Check. Output False-Friends-Liste Korrekturvorschlaege sichere Formulierungen. Abgrenzung zu common-law-begriffe-uebersetzung (Übersetzung) und common-law-bilingual-contract-review (Vergleich).
.claude/skills/thomasmoreai-common-law-false-friends-scanner/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 14% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 38% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 45% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-20 | ✓→✗ | ▼ Worse | 140% | 0% |
Bürgschaft, Garantie, Gewährleistung, Rücktritt, Vertretung, Schadensersatz, Eigentum, Besitz, Kündigung, Verzug und ähnliche Begriffe werden funktional statt wörtlich geprüft.
Common-Law-Kompass arbeitet freundlich, präzise und verzeihend. Der Stil darf leicht sein, aber nie auf Kosten der juristischen Trennschärfe.
| English Term | German False Friend | Correct Reading | Risk | |---|---|---|---| | "guarantee" | Garantie (absolute) | often secondary obligation (needs principal default) | HIGH | | "indemnity" | Entschaedigung / Buergschaft | primary independent obligation; survives variation | HIGH | | "warranty" | Gewaehrleistung | not Ruecktrittsrecht; damages only | MEDIUM | | "condition" | Bedingung (§ 158 BGB) | fundamental term; breach = right to terminate | HIGH | | "innominate term" | unklar | intermediate term; remedy depends on gravity of breach | MEDIUM | | "consideration" | Gegenleistung | legal concept requiring bargain; inadequacy acceptable | HIGH | | "estoppel" | Verwirkung (partial) | different types; does not extinguish right but precludes denial | MEDIUM | | "good faith" | Treu und Glauben | NOT implied in English commercial contracts | HIGH | | "specific performance" | Erfuellungsklage | equitable; discretionary; not available as of right | MEDIUM | | "discovery" | Akteneinsicht | much broader; both sides must produce relevant docs | HIGH | | "deed" | Urkunde | has legal effect without consideration | MEDIUM |
Before proceeding, clarify:
FALSE FRIENDS SCAN REPORT
Document: [DESCRIPTION] — Date: [DATE]
No. | Term Found | Location | Risk | Recommended Change
----|-----------|----------|------|-------------------
1 | [guarantee] | Clause 5.2 | HIGH — no discharge clause | Add: "primary indemnity" or add Holme v Brunskill waiver
2 | [good faith] | Clause 8 | HIGH — unenforceable English law | Remove or replace with specific obligations
3 | [condition] | Clause 3 | MEDIUM — correct usage but confirm | Confirm classification intended| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 30,199 | 19,505 | -35% | 1 | 1 | 0% | 5,372 | 4,455 | -17% | 0 | 0 | — |
case-02 | fail→fail | 23,643 | 21,218 | -10% | 1 | 1 | 0% | 3,972 | 4,911 | +24% | 0 | 0 | — |
case-03 | fail→pass | 28,678 | 23,180 | -19% | 1 | 1 | 0% | 4,638 | 5,270 | +14% | 0 | 0 | — |
case-04 | pass→pass | 13,964 | 23,354 | +67% | 1 | 1 | 0% | 2,552 | 3,905 | +53% | 0 | 0 | — |
case-05 | fail→pass | 14,192 | 14,672 | +3% | 1 | 1 | 0% | 2,278 | 3,140 | +38% | 0 | 0 | — |
case-06 | fail→fail | 20,372 | 14,721 | -28% | 1 | 1 | 0% | 2,484 | 3,680 | +48% | 0 | 0 | — |
case-07 | fail→pass | 14,328 | 18,206 | +27% | 1 | 1 | 0% | 2,664 | 3,874 | +45% | 0 | 0 | — |
case-08 | pass→pass | 21,716 | 14,340 | -34% | 1 | 1 | 0% | 2,506 | 3,621 | +44% | 0 | 0 | — |
case-09 | fail→fail | 16,235 | 21,959 | +35% | 1 | 1 | 0% | 2,699 | 4,590 | +70% | 0 | 0 | — |
case-10 | pass→pass | 15,357 | 18,706 | +22% | 1 | 1 | 0% | 2,682 | 4,159 | +55% | 0 | 0 | — |
case-11 | pass→pass | 17,002 | 19,234 | +13% | 1 | 1 | 0% | 2,500 | 4,247 | +70% | 0 | 0 | — |
case-12 | pass→pass | 18,246 | 16,742 | -8% | 1 | 1 | 0% | 2,996 | 4,086 | +36% | 0 | 0 | — |
case-13 | pass→pass | 13,185 | 17,836 | +35% | 1 | 1 | 0% | 2,283 | 3,949 | +73% | 0 | 0 | — |
case-14 | pass→pass | 18,725 | 16,885 | -10% | 1 | 1 | 0% | 2,837 | 3,987 | +41% | 0 | 0 | — |
case-15 | fail→pass | 17,253 | 13,305 | -23% | 1 | 1 | 0% | 2,532 | 2,710 | +7% | 0 | 0 | — |
case-16 | pass→pass | 14,329 | 13,369 | -7% | 1 | 1 | 0% | 2,374 | 3,310 | +39% | 0 | 0 | — |
case-17 | pass→pass | 17,416 | 22,773 | +31% | 1 | 1 | 0% | 3,040 | 5,119 | +68% | 0 | 0 | — |
case-18 | pass→pass | 11,221 | 14,375 | +28% | 1 | 1 | 0% | 1,971 | 3,514 | +78% | 0 | 0 | — |
case-19 | fail→fail | 10,976 | 7,540 | -31% | 1 | 1 | 0% | 1,797 | 2,224 | +24% | 0 | 0 | — |
case-20 | pass→fail | 9,155 | 15,673 | +71% | 1 | 1 | 0% | 1,688 | 4,052 | +140% | 0 | 0 | — |
case-21 | pass→pass | 7,492 | 8,324 | +11% | 1 | 1 | 0% | 1,658 | 2,604 | +57% | 0 | 0 | — |
case-22 | pass→pass | 14,571 | 15,995 | +10% | 1 | 1 | 0% | 2,523 | 3,982 | +58% | 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. 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.