Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Drafts enforceable End-User License Agreements for software licensors across desktop, mobile, SaaS, and cloud models. Covers click-wrap formation, IP ownership, liability limitations, data privacy compliance (GDPR/CCPA/COPPA), and export controls. Use when drafting software license agreements, app store terms, SaaS subscription agreements, or trial/freemium license terms.
.claude/skills/thomasmoreai-eula/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-21 | ✗→✓ | ▲ Improved | 104% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | 85% | 0% |
| case-07 | ✓→✓ | ＝ Same ✓ | 99% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 47% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 64% | 0% |
Drafts an enforceable EULA protecting licensor IP, limiting liability, and satisfying consumer protection standards across jurisdictions.
Gather before drafting:
| Term | Scope | |---|---| | Software | Product name, version, included modules | | Updates | Bug fixes / security patches (included) | | Upgrades | Major versions (may require separate fee) | | Authorized Users | Named / concurrent / site license scope | | Documentation | Manuals, API docs, specs | | Confidential Information | Source code, algorithms, benchmarks, designated materials |
| Category | Prohibited Conduct | |---|---| | Reverse engineering | Decompile, disassemble, derive source — except as required by law for interoperability (EU Directive 2009/24/EC) | | Derivative works | Modify, translate, adapt, or create based on Software | | Distribution | Sublicense, rent, lease, lend, transfer, service-bureau use | | Circumvention | Bypass license enforcement, DRM, or security features (17 U.S.C. § 1201 VERIFY]) | | Notices | Remove/alter copyright, trademark, or proprietary legends | | Competitive use | Benchmark for competitive analysis; publish results without consent | | Safety-critical | Aircraft nav, nuclear, life support, weapons | | Export | Export/re-export violating EAR, ITAR, or OFAC sanctions |
> MUST BE ALL-CAPS OR BOLD (conspicuousness requirement)
> MUST BE ALL-CAPS OR BOLD
Licensor → User (IP, if offered):
User → Licensor:
Include provisions for each applicable regime:
| Regime | Trigger | Key Obligations | |---|---|---| | GDPR | EU/EEA users | Lawful basis, data subject rights, 72-hr breach notice, SCCs, DPO if required | | CCPA/CPRA | CA users | Category disclosure, opt-out of sale/sharing, deletion/correction rights | | COPPA | Users under 13 | Verifiable parental consent before collection | | HIPAA | PHI for covered entities | BAA required; Privacy & Security Rule compliance | | GLBA | Financial data | Safeguards Rule compliance | | FERPA | Student records | Institutional consent; no unauthorized disclosure |
| Provision | Key Points | |---|---| | Entire Agreement | Supersedes all prior; no extra-contractual reliance | | Amendment | Post + 30-day notice; material paid-license changes require re-acceptance | | Severability | Reform to minimum extent; per-jurisdiction independence | | Waiver | Written and signed; no implied waiver | | Assignment | User cannot assign; licensor assigns freely (including M&A) | | Force Majeure | Acts of God, war, pandemic, infrastructure failure; termination right if >60–90 days | | No Partnership | Independent contractors; no agency/JV/franchise | | Third-Party Beneficiaries | None except third-party component licensors (IP) | | Notices | Email (confirmed) or certified mail; in-software posting for general notices | | Counterparts | Electronic signatures valid |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 22,954 | 28,942 | +26% | 1 | 1 | 0% | 4,402 | 8,131 | +85% | 0 | 0 | — |
case-07 | pass→pass | 11,457 | 8,841 | -23% | 1 | 1 | 0% | 1,953 | 3,885 | +99% | 0 | 0 | — |
case-02 | pass→pass | 29,414 | 31,050 | +6% | 1 | 1 | 0% | 5,872 | 8,645 | +47% | 0 | 0 | — |
case-03 | pass→pass | 14,927 | 14,683 | -2% | 1 | 1 | 0% | 3,580 | 5,856 | +64% | 0 | 0 | — |
case-04 | fail→fail | 30,720 | 32,319 | +5% | 1 | 1 | 0% | 5,742 | 8,716 | +52% | 0 | 0 | — |
case-05 | fail→fail | 23,464 | 31,841 | +36% | 1 | 1 | 0% | 4,582 | 8,703 | +90% | 0 | 0 | — |
case-06 | fail→fail | 28,672 | 31,581 | +10% | 1 | 1 | 0% | 6,231 | 8,699 | +40% | 0 | 0 | — |
case-08 | pass→pass | 12,584 | 12,582 | -0% | 1 | 1 | 0% | 1,815 | 4,635 | +155% | 0 | 0 | — |
case-09 | pass→pass | 11,242 | 8,850 | -21% | 1 | 1 | 0% | 2,146 | 3,970 | +85% | 0 | 0 | — |
case-10 | pass→pass | 10,452 | 9,756 | -7% | 1 | 1 | 0% | 1,986 | 4,205 | +112% | 0 | 0 | — |
case-11 | pass→pass | 15,679 | 14,895 | -5% | 1 | 1 | 0% | 2,741 | 5,327 | +94% | 0 | 0 | — |
case-12 | pass→pass | 14,161 | 17,603 | +24% | 1 | 1 | 0% | 2,353 | 5,698 | +142% | 0 | 0 | — |
case-13 | pass→pass | 19,357 | 14,592 | -25% | 1 | 1 | 0% | 3,427 | 5,315 | +55% | 0 | 0 | — |
case-14 | pass→pass | 12,380 | 14,254 | +15% | 1 | 1 | 0% | 2,075 | 4,702 | +127% | 0 | 0 | — |
case-15 | pass→pass | 13,967 | 11,978 | -14% | 1 | 1 | 0% | 2,443 | 4,910 | +101% | 0 | 0 | — |
case-16 | pass→pass | 7,592 | 10,812 | +42% | 1 | 1 | 0% | 1,349 | 4,434 | +229% | 0 | 0 | — |
case-17 | pass→pass | 8,806 | 9,799 | +11% | 1 | 1 | 0% | 1,630 | 4,237 | +160% | 0 | 0 | — |
case-18 | pass→pass | 14,504 | 16,861 | +16% | 1 | 1 | 0% | 2,526 | 5,319 | +111% | 0 | 0 | — |
case-19 | pass→pass | 11,629 | 14,614 | +26% | 1 | 1 | 0% | 2,052 | 5,043 | +146% | 0 | 0 | — |
case-20 | pass→pass | 12,862 | 12,796 | -1% | 1 | 1 | 0% | 2,062 | 4,549 | +121% | 0 | 0 | — |
case-21 | fail→pass | 14,195 | 10,011 | -29% | 1 | 1 | 0% | 2,046 | 4,173 | +104% | 0 | 0 | — |
case-22 | pass→pass | 10,656 | 9,402 | -12% | 1 | 1 | 0% | 1,744 | 4,009 | +130% | 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 +5 percentage points is the difference between those two pass rates over the 22 comparable cases.
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.