What problem does it solve? Static analyzers like Roslyn catch compiler-level issues, but they miss language-dependent defects and system-level implementation risks such as async hangs, fake-clock wait loops, connection churn, silent fallbacks, and broken context propagation. This Skill performs a structured, evidence-based C# code review across twelve review categories and produces findings with severity, evidence, and remediation. ## Core Features & Use Cases - Twelve review categories: Covers attribute activation/precondition mismatch, resource efficiency, operational resilience, synchronization, business input integrity, support lifecycle, error handling, time/culture, state/determinism, uncertainty escalation, contract/schema resilience, and traceability/context propagation. - Gate verdict routing: Emits a pre-CI verdict (blocked, needs-review, or proceed) with evidence and downgrade reasons, routing the diff without asserting correctness. - Structured handoffs: Routes security findings, XDDP trace gaps, and design-assumption issues to specialist skills instead of silently absorbing them. - Use Case: Point the Skill at a .NET solution before merging a pull request; it establishes the Roslyn baseline, reviews each active category (optionally via subagents), and returns a check item matrix plus a findings list with file-level evidence and remediation. ## Quick Start Ask the AI to run a C# review on your repository or solution path, optionally specifying scope filters and whether you want findings only or findings with fixes.
