Now liveThe Skillselion MCP - thousands of ranked skills, loaded into your agent mid-task. No install.Get it →

hylarucoder/hai-stack

20 skills205 installs5.5k starsGitHub

Install

npx skills add https://github.com/hylarucoder/hai-stack

Skills in this repo

1GejuGenerates a sharp direction judgment during design discussions, recommending the right target model rather than the smallest patch. A developer uses it to open the design space, challenge conservative proposals, and keep the bold call testable.60installs2Clean Code ReviewerReviews code against Clean Code principles across seven dimensions and outputs a severity-sorted findings report. A developer uses it to catch code smells and get behavior-preserving refactor suggestions before committing.13installs3GoudiPressure-tests an over-ambitious plan and returns a grounded verdict (go/shrink/pause/reject/validate-first) with the smallest proof step, an explicit cut list, and a stop rule. A developer uses it when a discussion has more vision than executable grounding and needs to be scoped down to a concrete first move.11installs4Hai IdeaReturns one clear verdict per idea with a dimension-by-dimension scorecard, a reframe, and the cheapest validation test. A developer uses it to decide whether to build, ship, kill, postpone, or reframe an idea or feature.11installs5Hai GoalProduces a written goal document to stop coding-while-watching, defining what should become true and how completion is verified. A developer uses it before starting meaningful work or to re-anchor an existing plan around a target.10installs6Hai Visual ReportTurns an idea, PRD, review, or plan into a presentable multi-section HTML report artifact and returns its path. A developer uses it to visualize analysis or a review as a webpage rather than plain Markdown.9installs7Hai ArchitectureProduces an architecture review or design-decision critique grounded in Ousterhout's A Philosophy of Software Design, with an architecture map, ranked file:line findings, why-not alternatives, and a red/blue adversarial check. A developer uses it to judge module boundaries, abstraction depth, dependency direction, and change amplification.8installs8Hai TddDrives development through a strict red-green-refactor TDD loop, requiring a failing test first for the right reason before the minimal implementation, then a refactor. A developer uses it to implement a feature, fix a bug, or protect behavior against regression with test-first discipline.8installs9Readme BeautifierReformats a messy README or similar markdown doc into a clean, consistently formatted version using a prioritized checklist for headings, lists, code blocks, tables, and spacing. A developer uses it when a project doc looks disorganized but its content should stay untouched.8installs10Entity Model AuditorProduces a structured data-model audit for each entity from a PRD and optional codebase, judging what fields should exist and where they should live. A developer uses it when designing a schema, checking PRD fields against a DB, or planning a migration.7installs11Hai PrdHandles all PRD work: whether one is needed, granularity, splitting/merging, drafting, and refining. A developer uses it to write, rewrite, audit, or scope a product requirements document.7installs12React Component DiagnosisRuns a deep health check on a single React component across seven dimensions and outputs a structured diagnosis with P0/P1/P2 recommendations. A developer uses it to review a component's design, props/API, or re-render behavior.7installs13Create Visual CardGenerates a decorative visual information card (Swiss-grid, bold type) as a self-contained HTML file and renders it to a PNG. A developer uses it to make one shareable card image such as a quote, summary, or social cover.6installs14Hai Ast GrepShips ast-grep patterns and rules that structurally search, lint, or rewrite code via tree-sitter, validated against positive and negative fixtures. A developer uses it for repo-wide codemods, call-site searches, or CI lint guards where grep over- or under-matches.6installs15Hai Audit Docs Against CodeFinds stale or implementation-inconsistent doc claims backed by evidence from source, config, and contract files, ranked P0-P3. A developer uses it to verify docs still match the code after a rename or refactor.6installs16Hai Audit Docs InternallyChecks docs for self-consistency without comparing to code, reporting P0-P3 findings plus update/move/merge/remove/split decisions. A developer uses it when a PRD or spec contradicts itself, repeats sections, or drifts in terminology.6installs17Hai NamingNames or renames concepts, variables, functions, types, and entities by treating naming as architecture expressed in language. A developer uses it when a name feels vague, inconsistent, or misleading, or when reviewing naming across code.6installs18Hai RazorApplies Occam's razor to a PRD, architecture, workflow, or data model, classifying every concept with evidence and a cut/proof list. A developer uses it to trim over-engineering, fake requirements, and premature abstractions.6installs19Hai Rewrite DocVerifies a patched-over, drifted document block by block against the codebase and current conclusions, keeps what passes verbatim, removes what fails, and ships a disposition table plus open-questions list. A developer uses it when a doc has accumulated doubtful claims and needs re-verification rather than more patching.5installs20Hai SsotHunts down single-source-of-truth violations across ten symptom classes (multi-source literals, shape proliferation, redundant conversions, re-implemented derivations, and more) and produces a numbered findings report with severity and treatment recipes. A developer uses it when they suspect duplicated definitions or drift between layers.5installs

This week in AI coding

Five minutes, every Monday - the tools, releases and tactics for developers.

unsubscribe anytime.

hylarucoder/hai-stack · 20 skills · Skillselion