Menu
SkillsMP a collecté 17 skills depuis KiemVM/AI-SDLC. Ouvrez un skill pour examiner sa source et ses détails.
Affichage de 17 skills collectés sur 17.
Use when running /decide, or whenever a product/technical decision changes something already written (a spec, plan, architecture doc) mid-work — captures it as an ADR and finds what needs to be updated downstream. Philosophy adapted from GitHub Spec-Kit's ADR…
Langue du texte source : anglais
Use when running /spec:plan, after a Specification.md exists and before task breakdown — turns approved requirements into a technical approach, file structure, and testing strategy. Philosophy adapted from GitHub Spec-Kit's plan phase and BMAD-METHOD's…
Langue du texte source : anglais
Use when running /engineering:review, or whenever asked to review a diff or PR for bugs and spec/constitution adherence. If the host project already has Anthropic's official code-review plugin configured, prefer that instead. Philosophy adapted from…
Langue du texte source : anglais
Use when running /spec:spec, or whenever turning a raw feature idea into a written specification — before any implementation planning or code. Philosophy adapted from GitHub Spec-Kit's spec-driven development approach.
Langue du texte source : anglais
Use when implementing any feature or bugfix, before writing any implementation code. Triggered by /engineering:implement and /engineering:test. Philosophy adapted from Superpowers and Matt Pocock's skills repo TDD discipline.
Langue du texte source : anglais
Use when running /business:vision, or whenever capturing a product's problem, target users, and success criteria at the highest level, before any PRD or feature work. Philosophy adapted from BMAD-METHOD's Analyst-stage discovery.
Langue du texte source : anglais
Use when running /business:architecture, after a PRD exists — documents the current-state whole-system architecture. Philosophy adapted from BMAD-METHOD's Architect-stage documentation-first approach.
Langue du texte source : anglais
Use when running /business:persona, or whenever defining a user persona that later UX work (journeys, wireframes) will reference. Philosophy adapted from BMAD-METHOD's discovery stage and the toolkit's UX extension.
Langue du texte source : anglais
Use when running /business:prd, after a Vision.md exists — derives a product-level epic list from the vision's goals. Philosophy adapted from BMAD-METHOD's PM-stage discovery.
Langue du texte source : anglais
Use when running /ux:prototype, after wireframes exist — stitches them into a flow and checks readiness for specification. Philosophy adapted from the toolkit's UX extension (Wireframe -> Prototype -> Specification).
Langue du texte source : anglais
Use when running /release:release — gathers completed features, checks release readiness, and drafts changelog/release notes. Never performs the release itself (no git tag/push/deploy). Philosophy adapted from BMAD-METHOD's governance discipline and this…
Langue du texte source : anglais
Use when running /ux:user-journey, after a persona exists — walks that persona through a scenario end to end. Philosophy adapted from the toolkit's UX extension (Business Goal -> Persona -> Journey -> Wireframe -> Prototype).
Langue du texte source : anglais
Use when running /ux:wireframe, or whenever laying out a screen's structure and content priority before visual design or implementation. Philosophy adapted from the toolkit's UX extension and low-fidelity wireframing practice.
Langue du texte source : anglais
Use when running /engineering:refactor, or when code needs structural improvement without changing behavior. Philosophy adapted from Matt Pocock's skills repo architecture-evolution approach.
Langue du texte source : anglais
Use when encountering any bug, test failure, or unexpected behavior — before proposing or attempting a fix. Triggered from /engineering:implement and /engineering:test on unexplained failures. Philosophy adapted from Superpowers' systematic debugging…
Langue du texte source : anglais
Use when running /spec:tasks, after an ImplementationPlan.md exists — decomposes it into a small, orderable, testable task backlog. Philosophy adapted from GitHub Spec-Kit's tasks phase and Superpowers' planning discipline.
Langue du texte source : anglais
Use before claiming any task, test run, fix, or review is complete/passing — requires fresh command output as evidence, not memory or assumption. Invoked at the tail of /engineering:implement, /engineering:test, and per-step inside /engineering:refactor.…
Langue du texte source : anglais