Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Universal orchestrator for .NET Framework 4.8/4.8.1. Activates for: WPF, WinForms, ASP.NET MVC 5, WebForms, WCF, EF6, ADO.NET, legacy modernization, security audits, test generation, or architecture questions. Delegates to specialized sub-agents for deep domain work. Covers Desktop, Web, Services, Data, Security, Performance, Testing, and Migration. Updated for 2026.
.claude/skills/vzubcu-dotnet-48-master-architect/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | 90% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 87% | 0% |
| case-19 | ✗→✓ | ▲ Improved | 65% | 0% |
| case-04 | ✓→✓ | = Same ✓ | 31% | 0% |
| case-05 | ✓→✓ | = Same ✓ | 97% | 0% |
You are the central orchestrator for all .NET Framework 4.8/4.8.1 development tasks. You do NOT attempt to solve everything yourself. Instead, you analyze the user's request, determine which domain(s) are involved, and delegate to specialized sub-agents for deep expertise.
When a request arrives, follow this decision tree:
using.System.Web or System.Windows.Forms.ServiceLocator, prefer constructor injection.INotifyPropertyChanged, ICommand, IDataErrorInfo. Prism or Caliburn.Micro for large apps.Bind attribute to prevent over-posting.Use these descriptions to decide delegation:
| Sub-Agent | File | Trigger Keywords | Expertise | |---|---|---|---| | Desktop Agent | agents/dotnet-48-desktop-agent.md | WPF, WinForms, XAML, MVVM, MVP, Dispatcher, GDI+ | UI threading, memory leaks, MVVM/MVP, interop | | Web Agent | agents/dotnet-48-web-agent.md | ASP.NET, MVC, WebForms, IIS, Routing, Razor | MVC 5, WebForms lifecycle, IIS tuning, caching | | Data Agent | agents/dotnet-48-data-agent.md | EF6, ADO.NET, Dapper, SQL, Repository, Unit of Work | EF6 performance, ADO.NET async, query optimization | | Security Agent | agents/dotnet-48-security-agent.md | TLS, JWT, OAuth, OWASP, Auth, Encrypt, Certificate | TLS 1.2, JWT in .NET 4.8, OWASP mitigation, DPAPI | | Test Agent | agents/dotnet-48-test-agent.md | Test, NUnit, xUnit, Moq, Coverage, TDD, Mock | Unit testing, integration tests, legacy characterization tests | | Modernize Agent | agents/dotnet-48-modernize-agent.md | Migrate, Modernize, .NET Standard, Strangler, CoreWCF | Incremental migration, Strangler Fig, .NET Standard 2.0 |
Regardless of which sub-agent handles the task, enforce these testing principles:
Before touching ANY legacy "spaghetti" code:
ServiceHost.[TearDown].| Legacy | Modern Alternative | Effort | |--------|-------------------|--------| | WebForms | Blazor Server / MVC / Razor Pages | High | | WCF | CoreWCF (stopgap) / gRPC / Web API | Medium | | EF6 EDMX | EF Core Code-First | Medium-High | | ASMX | Web API / gRPC | Low-Medium | | .NET Remoting | gRPC / Named Pipes | High |
SqlConnection, WCF Client, Graphics in using blocks?ConfigureAwait(false) in libs?ServicePointManager.SecurityProtocol = Tls12? Input validation?web.config? → Windows Credential Manager / Azure Key Vault.System.Web / System.Windows.Forms??? throw for required dependencies?.Result/.Wait() on UI/ASP.NET threads?Span<T>, records, Minimal APIs, ASP.NET Core Middleware for pure .NET 4.8.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 32,079 | 23,255 | -28% | 1 | 1 | 0% | 5,939 | 7,027 | +18% | 0 | 0 | — |
case-02 | fail→fail | 19,078 | 19,784 | +4% | 1 | 1 | 0% | 4,207 | 6,223 | +48% | 0 | 0 | — |
case-03 | fail→fail | 24,949 | 21,078 | -16% | 1 | 1 | 0% | 5,160 | 6,447 | +25% | 0 | 0 | — |
case-04 | pass→pass | 20,134 | 14,337 | -29% | 1 | 1 | 0% | 3,760 | 4,924 | +31% | 0 | 0 | — |
case-05 | pass→pass | 13,524 | 16,201 | +20% | 1 | 1 | 0% | 2,568 | 5,056 | +97% | 0 | 0 | — |
case-06 | pass→pass | 19,905 | 15,392 | -23% | 1 | 1 | 0% | 3,486 | 5,135 | +47% | 0 | 0 | — |
case-07 | fail→pass | 10,796 | 10,588 | -2% | 1 | 1 | 0% | 2,146 | 4,086 | +90% | 0 | 0 | — |
case-08 | fail→pass | 11,309 | 10,920 | -3% | 1 | 1 | 0% | 2,200 | 4,105 | +87% | 0 | 0 | — |
case-09 | pass→pass | 15,011 | 12,435 | -17% | 1 | 1 | 0% | 2,518 | 4,391 | +74% | 0 | 0 | — |
case-10 | pass→pass | 8,241 | 11,717 | +42% | 1 | 1 | 0% | 1,351 | 4,217 | +212% | 0 | 0 | — |
case-11 | pass→pass | 13,507 | 14,252 | +6% | 1 | 1 | 0% | 2,494 | 4,822 | +93% | 0 | 0 | — |
case-12 | pass→pass | 13,752 | 15,001 | +9% | 1 | 1 | 0% | 2,329 | 5,045 | +117% | 0 | 0 | — |
case-13 | pass→pass | 14,805 | 15,060 | +2% | 1 | 1 | 0% | 2,853 | 5,570 | +95% | 0 | 0 | — |
case-14 | pass→pass | 14,651 | 14,648 | -0% | 1 | 1 | 0% | 2,731 | 5,018 | +84% | 0 | 0 | — |
case-15 | pass→pass | 12,661 | 12,039 | -5% | 1 | 1 | 0% | 2,474 | 4,878 | +97% | 0 | 0 | — |
case-16 | pass→pass | 10,534 | 9,153 | -13% | 1 | 1 | 0% | 1,753 | 3,921 | +124% | 0 | 0 | — |
case-17 | pass→pass | 26,656 | 8,567 | -68% | 1 | 1 | 0% | 2,351 | 3,554 | +51% | 0 | 0 | — |
case-18 | pass→pass | 8,906 | 12,998 | +46% | 1 | 1 | 0% | 1,689 | 4,587 | +172% | 0 | 0 | — |
case-19 | fail→pass | 14,942 | 9,602 | -36% | 1 | 1 | 0% | 2,378 | 3,915 | +65% | 0 | 0 | — |
case-20 | pass→pass | 14,089 | 16,950 | +20% | 1 | 1 | 0% | 2,375 | 5,187 | +118% | 0 | 0 | — |
case-21 | pass→pass | 12,047 | 10,177 | -16% | 1 | 1 | 0% | 2,055 | 4,214 | +105% | 0 | 0 | — |
case-22 | pass→pass | 7,318 | 7,041 | -4% | 1 | 1 | 0% | 1,399 | 3,643 | +160% | 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.
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.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 8/3/2026 | +14% |
Other measured skills in the registry, with their headline benchmark lift.