Use when the user wants to find and remove code duplication, says "deduplicate", "art-dupl", or wants to reduce duplication to ZERO. This skill is all about JUDGMENT and ZERO HARMFUL duplication. It does NOT mean zero report lines.
Splits or merges Go modules, refining go.mod boundaries. Direction-neutral. Use when the user wants to modularize a Go project, create independently versioned Go packages, restructure a Go project, or says "modularize", "split modules", "sub-modules", "go.mod per package", "multi-module", "module boundaries", "dependency isolation", "Go workspace", "go.work", "CQRS modularization", "break apart Go project", "too coupled", "Go package boundaries", "remodularize", "re-modularize", "merge modules", "too many modules", "over-modularized", "consolidate modules", "unix-style", "decompose deeper", "too few modules", "god-module", "composable modules".
Reviews the current architecture for scalability, modularity, service orientation, and composability. Use when the user asks about architecture quality, scalability, modularity, how to make the codebase more service-oriented or composable, or says "architecture review".
Triggers a brutally honest self-review of recent work and the current codebase state (Go projects). Use when the user asks for self-reflection, self-critique, wants to know what was forgotten, what's stupid, what could be better, find ghost systems, split brains, or wants a comprehensive improvement plan with Go ecosystem library awareness.
Runs build, lint, and code duplication analysis to surface all code quality issues. Use when the user asks to check code quality, run build and lint, find code duplication, or wants a sorted list of all issues.
Reviews and redesigns data models from first principles using Go's type system. Use when the user wants to review a data model, redesign types, improve a domain model, fix stringly-typed code, add type safety, use Go generics, composition, or interfaces, or says "data model", "review types", "redesign schema", "type system", "make impossible states unrepresentable", "branded types", "domain model", "entity design", "data architecture", "type safety review", "refactor data layer", "model redesign", "Go data model", "Go types", "Go generics", "Go structs", "Go interfaces", "Go composition". Triggers for any language but produces Go as the primary design language. Also fires when the user asks about validation layers, schema design, or converting implicit models to explicit typed ones.
Performs a comprehensive code review visiting every single code and test file as a Senior Software Architect. Use when the user wants a full codebase review, every file reviewed, deep quality audit, or says "visit every file", "review everything", "full review". Checks type safety, architecture, split brains, duplications, and fixes or adds TODOs on the spot. Includes Pareto-based planning and prioritization.
Performs a deep-dive research audit on a single library to determine whether the project uses it to its FULL potential. Use this skill when the user asks "are we using X to the max", "are we fully leveraging <library>", "deep dive into <library>", "library utilization audit", "library adoption review", "library maximization", "are we using <library> correctly", "what features of <library> are we missing", "library capabilities audit", "library usage review", "do we use <library> to its full potential", or any question about whether a specific dependency is underutilized, misused, or behind on features. Also triggers when the user names a library (e.g. Prisma, Zod, GORM, sqlc, Effect, RxJS, gorilla/mux, cobra, viper, react-query, zod, drizzle, typeorm, ent, elm, dayjs, lodash, immer, zustand, effector) and wants to know if they're getting the most out of it. Fires for ANY language and ANY library — JS/TS, Go, Rust, Python, Ruby, etc.