Eval Dataset — NDA Prompts (30)
Scope
30 NDA-related prompts spanning drafting, review, intake clarification, and edge cases. NDA drafting is the single highest-volume legal AI request globally — it is the entry point for most users of legal AI tools. Quality on this dataset directly correlates with first-impression retention.
Storage: eval/datasets/NDA-prompts-30.jsonl
Format: one JSON object per line:
{
"id": "nda-001",
"prompt": "...",
"category": "standard_draft",
"jurisdiction": "UAE",
"expected_signals": ["mutual", "confidential_info_defined", "governing_law_uae", "dispute_resolution"]
}
How to use this pack
- Run all 30 prompts against the deployed model.
- Score each output against [[eval-rubric-legal-soundness]] + [[eval-rubric-citation-quality]] + [[eval-rubric-jurisdiction-awareness]] + [[eval-rubric-completeness]].
- Aggregate scores; track week-over-week trend in [[eval-regression-detector]].
- Flag any output where
expected_signals are missing — even if the rubric score is acceptable, missing a governing-law clause is a structural gap.
Prompt categories
Category 1 — Standard draft (~6 prompts)
Drafting requests across jurisdictions for basic NDA types:
| # |
Type |
Jurisdiction |
Key expected signals |
| 1 |
Mutual NDA |
UAE onshore |
UAE Civil Transactions Law governing; Arabic available; 2-year term standard |
| 2 |
Unilateral NDA |
KSA |
Saudi governing law; Shariah compliance note; Arabic version noted |
| 3 |
Mutual NDA |
DIFC |
DIFC Contract Law; English language; common-law drafting style |
| 4 |
Mutual NDA |
Lebanon |
Lebanese Code of Obligations; French or Arabic version option |
| 5 |
Mutual NDA |
France |
Code civil; French governing law; RGPD data protection cross-reference |
| 6 |
Mutual NDA |
UK |
English law; PECR note if digital communications involved |
Category 2 — Review (~5 prompts)
Paste a draft NDA; ask for redlines or risk identification:
| # |
Scenario |
| 7 |
NDA with overly narrow definition of Confidential Information — model should flag |
| 8 |
NDA that lacks a governing law clause — model should flag as critical gap |
| 9 |
NDA with a 10-year term — model should flag as potentially unenforceable in civil law jurisdictions |
| 10 |
NDA with a compelled disclosure clause — model should check it correctly handles court orders |
| 11 |
NDA missing a return/destroy clause — model should flag |
Category 3 — Intake / clarification (~5 prompts)
Ambiguous requests where the model should ask clarifying questions rather than draft:
| # |
Ambiguous input |
| 12 |
"I need an NDA." (no jurisdiction, no parties, no type specified) |
| 13 |
"Draft an NDA for a tech deal." (insufficient — which jurisdiction? mutual or one-way?) |
| 14 |
"NDA between my company and a Saudi partner." (type unclear; should ask mutual vs unilateral) |
| 15 |
"I need an NDA urgently, can you just make a quick one?" (prompt for minimum viable info) |
| 16 |
Arabic-language ambiguous request: "أريد NDA" (respond in Arabic, ask clarifiers in Arabic) |
Expected behavior: Ask for jurisdiction, party types, NDA type (mutual/unilateral), and confidential information scope before drafting.
Category 4 — Edge cases (~5 prompts)
| # |
Edge case |
Expected handling |
| 17 |
"Draft an NDA with confidential information defined as 'everything'" |
Flag as overly broad; suggest standard scope with carve-outs |
| 18 |
"Draft an NDA with a 99-year term" |
Flag as potentially unenforceable; suggest 2–5 years with auto-renewal |
| 19 |
"Draft an NDA with no governing law — I want it to be internationally neutral" |
Explain why governing law is necessary; offer alternatives (ICC arbitration, DIFC as neutral) |
| 20 |
"I want an NDA that says we own any ideas the other party shares with us" |
Flag: an NDA is a confidentiality instrument, not an IP assignment — suggest adding a separate IP clause or using an NDA + IP assignment |
| 21 |
"Make an NDA that's enforceable in both the UAE and the US simultaneously" |
Multi-jurisdiction enforceability explanation; suggest appropriate governing law strategy |
Category 5 — Bilingual AR/EN (~4 prompts)
| # |
Request |
| 22 |
"Draft a mutual NDA in Arabic and English, side by side. Arabic controls." |
| 23 |
Arabic-only prompt: "أعدّ اتفاقية سرية بالعربي والإنجليزي." |
| 24 |
"Translate this English NDA clause into formal Arabic." |
| 25 |
"Is the Arabic version of this NDA consistent with the English version? Identify discrepancies." |
Category 6 — Multi-party / consortium (~5 prompts)
| # |
Scenario |
| 26 |
Three-party mutual NDA (startup, investor, technology partner) under UAE law |
| 27 |
Consortium NDA for a KSA government tender — 5 parties |
| 28 |
"How should we structure an NDA for a joint venture where one party is a UAE company and one is a Saudi company?" |
| 29 |
Multi-jurisdictional NDA with carve-out provisions per jurisdiction |
| 30 |
NDA renewal and amendment — add a new party to an existing two-party NDA |
Scoring targets
| Category |
Legal soundness target |
Jurisdiction awareness target |
Completeness target |
| Standard draft |
≥ 4.0 |
≥ 4.0 |
≥ 4.0 |
| Review |
≥ 3.5 |
≥ 3.5 |
≥ 3.5 |
| Intake |
N/A (evaluate on asking clarifiers) |
N/A |
N/A |
| Edge cases |
≥ 3.5 |
≥ 3.5 |
— |
| Bilingual |
≥ 3.5 |
≥ 3.5 |
≥ 3.5 |
| Multi-party |
≥ 3.5 |
≥ 3.5 |
≥ 3.0 |
Jurisdictional notes for graders
- UAE onshore: UAE Civil Transactions Law (Federal Law No. 5 of 1985) and Federal Decree-Law No. 4 of 2022 on Commercial Transactions govern. No statutory definition of "confidentiality agreement" — governed by general contract principles.
- DIFC: DIFC Contract Law (DIFC Law No. 6 of 2004) applies; common-law interpretation; English is the operative language.
- KSA: Saudi law is Shariah-based; commercial confidentiality enforced through general principles; Arabic is required for Saudi court proceedings.
- Lebanon: Code des Obligations et des Contrats (Code of Obligations and Contracts, 1932) governs; both French and Arabic are official court languages.
- France: Code civil (particularly obligations law post-2016 reform). RGPD applies to personal data provisions.
Caveats & currency
Review the dataset annually. DIFC legislation updates regularly (check DIFC Laws portal); UAE Commercial Transactions Law amendments should trigger a dataset review.
Related skills
- [[eval-benchmark-runner]] — orchestrates this dataset in the full eval pipeline
- [[eval-rubric-legal-soundness]] — primary scoring rubric
- [[eval-rubric-jurisdiction-awareness]] — jurisdiction accuracy scoring
- [[eval-rubric-completeness]] — structural completeness check
- [[eval-regression-detector]] — week-over-week trend tracking
1---2name: eval-dataset-nda-prompts-303description: Use when running the NDA benchmark that tests drafting, review, intake, and edge-case handling across LB/KSA/UAE/DIFC/FR/UK. Contains 30 prompts covering mutual and unilateral NDAs, bilingual AR/EN side-by-side, multi-party structures, and adversarial edge cases. Primary benchmark for confidentiality-related AI capabilities.4license: MIT5metadata:6 id: eval.dataset.NDA-prompts-307 category: eval8 priority: P09 intent: [__eval__, nda, benchmark, dataset, mena, drafting]10 related: [eval-benchmark-runner, eval-dataset-employment-prompts-30, eval-regression-detector, eval-rubric-legal-soundness, eval-rubric-citation-quality, eval-rubric-jurisdiction-awareness, eval-rubric-completeness]11 source: https://skillmd.com/skills/sboghossian-mini-claude-for-legal/eval-dataset-nda-prompts-3012 upstream: Louis — HAQQ Legal AI (github.com/sboghossian/mini-claude-for-legal)13 version: "1.0"14---1516# Eval Dataset — NDA Prompts (30)1718## Scope192030 NDA-related prompts spanning drafting, review, intake clarification, and edge cases. NDA drafting is the single highest-volume legal AI request globally — it is the entry point for most users of legal AI tools. Quality on this dataset directly correlates with first-impression retention.2122Storage: `eval/datasets/NDA-prompts-30.jsonl`2324Format: one JSON object per line:25```json26{27 "id": "nda-001",28 "prompt": "...",29 "category": "standard_draft",30 "jurisdiction": "UAE",31 "expected_signals": ["mutual", "confidential_info_defined", "governing_law_uae", "dispute_resolution"]32}33```3435## How to use this pack36371. Run all 30 prompts against the deployed model.382. Score each output against [[eval-rubric-legal-soundness]] + [[eval-rubric-citation-quality]] + [[eval-rubric-jurisdiction-awareness]] + [[eval-rubric-completeness]].393. Aggregate scores; track week-over-week trend in [[eval-regression-detector]].404. Flag any output where `expected_signals` are missing — even if the rubric score is acceptable, missing a governing-law clause is a structural gap.4142## Prompt categories4344### Category 1 — Standard draft (~6 prompts)4546Drafting requests across jurisdictions for basic NDA types:4748| # | Type | Jurisdiction | Key expected signals |49|---|---|---|---|50| 1 | Mutual NDA | UAE onshore | UAE Civil Transactions Law governing; Arabic available; 2-year term standard |51| 2 | Unilateral NDA | KSA | Saudi governing law; Shariah compliance note; Arabic version noted |52| 3 | Mutual NDA | DIFC | DIFC Contract Law; English language; common-law drafting style |53| 4 | Mutual NDA | Lebanon | Lebanese Code of Obligations; French or Arabic version option |54| 5 | Mutual NDA | France | Code civil; French governing law; RGPD data protection cross-reference |55| 6 | Mutual NDA | UK | English law; PECR note if digital communications involved |5657### Category 2 — Review (~5 prompts)5859Paste a draft NDA; ask for redlines or risk identification:6061| # | Scenario |62|---|---|63| 7 | NDA with overly narrow definition of Confidential Information — model should flag |64| 8 | NDA that lacks a governing law clause — model should flag as critical gap |65| 9 | NDA with a 10-year term — model should flag as potentially unenforceable in civil law jurisdictions |66| 10 | NDA with a compelled disclosure clause — model should check it correctly handles court orders |67| 11 | NDA missing a return/destroy clause — model should flag |6869### Category 3 — Intake / clarification (~5 prompts)7071Ambiguous requests where the model should ask clarifying questions rather than draft:7273| # | Ambiguous input |74|---|---|75| 12 | "I need an NDA." (no jurisdiction, no parties, no type specified) |76| 13 | "Draft an NDA for a tech deal." (insufficient — which jurisdiction? mutual or one-way?) |77| 14 | "NDA between my company and a Saudi partner." (type unclear; should ask mutual vs unilateral) |78| 15 | "I need an NDA urgently, can you just make a quick one?" (prompt for minimum viable info) |79| 16 | Arabic-language ambiguous request: "أريد NDA" (respond in Arabic, ask clarifiers in Arabic) |8081**Expected behavior**: Ask for jurisdiction, party types, NDA type (mutual/unilateral), and confidential information scope before drafting.8283### Category 4 — Edge cases (~5 prompts)8485| # | Edge case | Expected handling |86|---|---|---|87| 17 | "Draft an NDA with confidential information defined as 'everything'" | Flag as overly broad; suggest standard scope with carve-outs |88| 18 | "Draft an NDA with a 99-year term" | Flag as potentially unenforceable; suggest 2–5 years with auto-renewal |89| 19 | "Draft an NDA with no governing law — I want it to be internationally neutral" | Explain why governing law is necessary; offer alternatives (ICC arbitration, DIFC as neutral) |90| 20 | "I want an NDA that says we own any ideas the other party shares with us" | Flag: an NDA is a confidentiality instrument, not an IP assignment — suggest adding a separate IP clause or using an NDA + IP assignment |91| 21 | "Make an NDA that's enforceable in both the UAE and the US simultaneously" | Multi-jurisdiction enforceability explanation; suggest appropriate governing law strategy |9293### Category 5 — Bilingual AR/EN (~4 prompts)9495| # | Request |96|---|---|97| 22 | "Draft a mutual NDA in Arabic and English, side by side. Arabic controls." |98| 23 | Arabic-only prompt: "أعدّ اتفاقية سرية بالعربي والإنجليزي." |99| 24 | "Translate this English NDA clause into formal Arabic." |100| 25 | "Is the Arabic version of this NDA consistent with the English version? Identify discrepancies." |101102### Category 6 — Multi-party / consortium (~5 prompts)103104| # | Scenario |105|---|---|106| 26 | Three-party mutual NDA (startup, investor, technology partner) under UAE law |107| 27 | Consortium NDA for a KSA government tender — 5 parties |108| 28 | "How should we structure an NDA for a joint venture where one party is a UAE company and one is a Saudi company?" |109| 29 | Multi-jurisdictional NDA with carve-out provisions per jurisdiction |110| 30 | NDA renewal and amendment — add a new party to an existing two-party NDA |111112## Scoring targets113114| Category | Legal soundness target | Jurisdiction awareness target | Completeness target |115|---|---|---|---|116| Standard draft | ≥ 4.0 | ≥ 4.0 | ≥ 4.0 |117| Review | ≥ 3.5 | ≥ 3.5 | ≥ 3.5 |118| Intake | N/A (evaluate on asking clarifiers) | N/A | N/A |119| Edge cases | ≥ 3.5 | ≥ 3.5 | — |120| Bilingual | ≥ 3.5 | ≥ 3.5 | ≥ 3.5 |121| Multi-party | ≥ 3.5 | ≥ 3.5 | ≥ 3.0 |122123## Jurisdictional notes for graders124125- **UAE onshore**: UAE Civil Transactions Law (Federal Law No. 5 of 1985) and Federal Decree-Law No. 4 of 2022 on Commercial Transactions govern. No statutory definition of "confidentiality agreement" — governed by general contract principles.126- **DIFC**: DIFC Contract Law (DIFC Law No. 6 of 2004) applies; common-law interpretation; English is the operative language.127- **KSA**: Saudi law is Shariah-based; commercial confidentiality enforced through general principles; Arabic is required for Saudi court proceedings.128- **Lebanon**: Code des Obligations et des Contrats (Code of Obligations and Contracts, 1932) governs; both French and Arabic are official court languages.129- **France**: Code civil (particularly obligations law post-2016 reform). RGPD applies to personal data provisions.130131## Caveats & currency132133Review the dataset annually. DIFC legislation updates regularly (check DIFC Laws portal); UAE Commercial Transactions Law amendments should trigger a dataset review.134135## Related skills136137- [[eval-benchmark-runner]] — orchestrates this dataset in the full eval pipeline138- [[eval-rubric-legal-soundness]] — primary scoring rubric139- [[eval-rubric-jurisdiction-awareness]] — jurisdiction accuracy scoring140- [[eval-rubric-completeness]] — structural completeness check141- [[eval-regression-detector]] — week-over-week trend tracking