All authors
Claude Skills by baov
github.com/baov11 skills0 installs0 views
- Ai Code RemediationCold audit of a codebase written mostly by AI, followed by a remediation plan prioritized under a strict doctrine (TDD, Mikado method, micro-iteration). Use it SYSTEMATICALLY when the user asks to "audit this codebase", "size up the debt in this AI-generated project", "clean up a vibe-coded project", "figure out what the agents produced", "remediation plan", "clean up this code", or doubts the quality of a project written largely by agents — even when the word "audit" never comes up. Covers 8...Votes: 0GitHub stars: 2
- Behavior Driven TestingTest strategy driven by behaviors rather than "one test per class". Use as soon as tests are written, designed or reviewed, when test strategy, coverage, mutation testing, mutation score or TDD come up, when someone asks "how do I test this", or complains about brittle tests. Use it also when someone proposes a "1 test ↔ 1 class" mapping, sets a numeric coverage target as the goal, or wonders whether their tests are worth anything — that is the reflex this skill corrects.Votes: 0GitHub stars: 2
- Clarify With ChoicesDoctrine for validating with multiple-choice questions — how a skill questions the user before writing anything, and how it degrades when the host agent offers no multiple-choice mechanism. Loaded by the skills that validate structuring decisions (codebase-cartographer, codebase-harness, ddd-advisor, product-strategy), rarely invoked directly. Use it too when writing or revising a skill that has to ask the user for a decision, or when wondering how many questions to ask, in what shape, or wha...Votes: 0GitHub stars: 2
- Code Assimilation QuizPost-implementation assimilation quiz — a teaching code review run as multiple-choice questions, one at a time, so the developer takes ownership of the code the AI wrote. Use ONLY when the user explicitly asks for it, never automatically. Typical triggers — "run the live code review", "quiz me on what was implemented", "check that I understood the code", "multiple-choice questions on the diff", "help me assimilate what you coded", "teaching code review". Do not confuse with a standard code re...Votes: 0GitHub stars: 2
- Codebase CartographerMaps an existing or brand-new codebase to produce structured business + technical documentation. Creates a `docs/` folder split into `docs/business/` (glossary, core features, test cases — one .md per case, grouped by feature) and `docs/technical/` (architecture, stack, test strategy, ADRs backed by git history analysis), validates every section with the user through multiple-choice questions, refactors to remove redundancy, and links it all from `AGENTS.md`. Use AS SOON AS the user asks to "...Votes: 0GitHub stars: 2
- Codebase HarnessDeterministic enforcement layer that keeps a codebase coherent while AI agents contribute to it. Four optional blocks offered through multiple-choice questions: (A) executable invariants — ADRs, architecture rules and shape thresholds (size, complexity) turned into custom pre-commit/CI linters; (B) test-case bridge to real tests — YAML front-matter + a coverage script; (D) mutation testing — checks that unreviewed tests would really catch a bug; tool detected from the stack, scope calibrated ...Votes: 0GitHub stars: 2
- Ddd AdvisorDDD audit of an existing codebase AND Domain-Driven Design guidance while development is under way. Use as soon as the user mentions DDD, an "anemic domain model", "aggregates", "entities vs value objects", "bounded context", "ubiquitous language", "context map", "anti-corruption layer", or asks a domain modeling question ("where does this domain logic belong?", "is this an entity?", "how do I split this monolith into contexts?"). Use it ALSO when an audit or a remediation surfaces scattered ...Votes: 0GitHub stars: 2
- Plan Driven DevDisciplined workflow for implementing a new feature OR fixing a bug in an existing code project. Forces a context-gathering phase before anything is written, produces a persistent plan validated step by step, applies behavior-driven TDD, and captures lessons learned. Use this skill whenever a user prompt looks like a code task in an existing project (keywords such as "implement", "add", "fix", "bug", "feature", "refactor", "change", "it should", or any multi-file code change request). The ski...Votes: 0GitHub stars: 2
- Premerge ReviewDefect-oriented pre-merge code review — first runs the branch through a gauntlet of mechanical checks (tests, lint, harness invariants, test-case bridge, mutation on the diff), then analyzes the diff against main at a depth calibrated on criticality, and produces a structured report with a GO/NO-GO verdict. A red gauntlet stops the review: no point hand-reading a diff that is about to change. Use SYSTEMATICALLY when the user asks for a "code review", "review my branch", "check my diff before ...Votes: 0GitHub stars: 2
- Product StrategyProduct strategy from an idea or an existing product: market watch (competitors, blue or red ocean), vision, roadmap, story mapping, a backlog of epics and features prioritized, and a PRD for one chosen feature. Use as soon as the user wants to "define the product vision", "build a roadmap", "do a story map", "fill or prioritize the backlog", "study the market or the competition", "find a blue ocean", "write a PRD", or starts from an idea with no code yet and asks what to build first. Works a...Votes: 0GitHub stars: 2
- Systematic DebuggingRigorous bug diagnosis methodology — find and prove the root cause before any fix. Use SYSTEMATICALLY whenever unexpected behavior has to be explained: "why isn't this working", "weird bug", "it crashes", "regression", "flaky behavior", "it worked before", "error I don't understand", "investigate", "diagnose", or any fix request whose cause is not yet identified and proven. Covers the DIAGNOSIS phase only: minimal reproduction, falsifiable hypotheses, bisection, targeted instrumentation, proo...Votes: 0GitHub stars: 2