Design or debug machine-code lifting into LLVM IR for analysis and recompilation. Use when the lifting pipeline is the subject, not for general binary reverse engineering.
Review existing code or diffs for maintainability, design decay, and architecture risks, reporting actionable Symptom → Source → Consequence → Remedy findings. Do not use for writing code from scratch.
Build NeverC or other LLVM compilers across frontend, IR, optimization, code generation, ABI, and target support. Use when the compiler pipeline is the subject; use narrower LLVM skills for isolated tooling or analysis.
Reverse compiled, packed, obfuscated, virtualized, or anti-analysis targets for an authorized CTF challenge. Use while understanding the target; switch to an exploitation skill once the vulnerability is known.
Build dynamic instrumentation, runtime tracing, coverage, profiling, or monitoring with LLVM-related tooling. Use static-analysis when execution is not part of the analysis.
Diagnose NeverC runtime tests that fail only or intermittently in CI. Use for platform-specific crashes, signals, nondeterminism, or local/runner divergence; not ordinary build errors or workflow configuration failures.
Teach LLVM, Clang, and compiler concepts or assemble a learning path. Use for explanations and educational resources, not implementation work covered by a specialized LLVM skill.
Analyze or implement LLVM code-protection transformations such as control-flow flattening, string protection, and virtualization. Use when the obfuscation pass or pipeline is the subject.
Implement or analyze LLVM optimization passes, pipelines, and generated-code quality. Use for IR transformation and pass behavior, including NeverC compiler performance, not general application tuning.
Apply or develop LLVM sanitizers, compiler hardening, and exploit mitigations. Use when secure compilation is central, not for general vulnerability analysis unrelated to the toolchain.
Build Clang plugins, LibTooling tools, LLDB extensions, or clangd/LSP integrations. Use for developer tooling around LLVM, not compiler pipeline implementation.
Build MLIR or CIR dialects, conversions, passes, and multi-level compiler pipelines. Use when MLIR/CIR is the implementation substrate, including ML and domain-specific compilers.
Implement LLVM-based dataflow, pointer, taint, dependency, or verification analyses. Use for static program-analysis algorithms and tools, not runtime instrumentation.
Diagnose reproducible bugs and test failures by gathering evidence, isolating the failing boundary, and testing a root-cause hypothesis. Use when the cause is unclear or an attempted fix did not work.
Build a feature or fix through short red-green-refactor cycles. Use when the user explicitly requests TDD, test-first development, or a red-green-refactor workflow.
Turn a product idea into a usable first release across scoping, implementation, polish, and handoff. Use for end-to-end MVP or launch work, not isolated feature requests.
Write an implementation plan from an established specification. Use when the user asks for a plan or when complex work needs a durable handoff before implementation.
Idioma do texto original: inglês
atualizado
Mostrando 17 de 17 skills coletadas.
Instalar com um assistente de IA
Copie este prompt para o assistente de IA que você usa.
Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/NeverSight/NeverC.