AI & ML

antoinebou12 Skill

Step-by-step and reflective problem-solving for complex or multi-step tasks. Use when tackling complex tasks, planning, design, analysis that may need revision, unclear scope, or when the user asks for step-by-step reasoning. When the sequential-thinking MCP tool is available, use it for hard problems; otherwise apply the same principles in plain reasoning.

antoinebou12 Skill

Create and validate diagrams with the uml-mcp MCP server (generate_uml, generate_uml_image, validate_uml, list_diagram_types, generate_uml_batch). Use for UML, Mermaid, D2, Graphviz, Kroki URLs, inline chat images, goat/umlet, or diagram_type questions.

ether-moon Skill

Improves user-facing Korean prose by making grammar, referents, and logical relationships explicit while keeping responses concise. Use whenever the expected conversational or requested output language is Korean, alongside any task-specific skill; do not use solely because a prompt contains Korean when the requested output is non-Korean, code-only, an exact quotation, or governed by another artifact language policy.

ether-moon Bundle

Verifies proposed agent-directive additions and audits existing CLAUDE.md, AGENTS.md, and referenced instruction files for bloat, duplication, ambiguity, and placement. Use when adding, changing, or reviewing agent directives or when an agent proposes modifying them.

ether-moon Skill

Delegates technical review to independent peer agents, validates their findings, and routes actionable results through bounded autofixing and decision escalation. Use when the user explicitly asks Codex, Claude, Gemini, another model, peer LLMs, or a second opinion to review or validate work; do not use for an ordinary single-agent review.

jarvusinnovations Bundle

Packaging-and-distribution recipe for baking an AXI CLI (axi-sdk-js) into a Claude Code skill — bundling it under the skill's scripts/ as a self-contained committed .mjs shipped via `npx skills add`. This is the build/ship companion to the upstream `axi` skill: `axi` covers what good agent-facing CLI output looks like (TOON, schemas, content-first); this covers how to ship that CLI *inside a skill*. Use whenever you're turning an AXI tool into an installable skill, or working on any of its moving parts — the esbuild → committed .mjs build, the bash shim, the SessionStart hook (install/uninstall/status, project vs global scope), generating SKILL.md from a single-source command reference, the CI drift gate, or splitting a lean every-session "home" view from an on-demand live "dashboard". Triggers: "bundle my CLI in a skill", "AXI in a skill", "ship this CLI via npx skills", "self-contained skill bundle", "add a session-start hook to my skill", "generate SKILL.md from the CLI", or noticing a repo with `scripts/b

jarvusinnovations Bundle

Jarvus house conventions for writing, testing, and linting dbt projects — model layering and grain, what each stage may/may not do, the no-inner-joins and semantic-alias rules, the fast (<5 min) local dev-loop requirement, the no-unvalidated-demo-work-in-main rule, the generic + quality-model + unit-test patterns, the sqlfluff config, and a credential-free multi-tenant CI gate. Use when building or reviewing dbt models, deciding materialization or test coverage, setting up dbt linting/CI, or when "dbt", "sqlfluff", "staging/intermediate/marts", "dbt test", or "TIDES" come up. This is the opinionated house layer ON TOP OF dbt-labs' first-party dbt skills (which cover the mechanics) — see "Relationship to other skills".

jarvusinnovations Bundle

Maintain, stabilize, and clean up the Jarvus/Slate ExtJS 6.2 classic-toolkit web apps — SlateAdmin, the Slate CBL apps (SlateTasksTeacher/Student/Manager, SlateDemonstrations*, SlateStudentCompetenciesAdmin), and anything else living in a sencha-workspace/. Use this whenever the task touches Ext.define, Ext.app.Controller, a Sencha workspace, app.json, an ExtJS view/store/model/proxy, jarvus-* or slate-* Sencha packages, writing or debugging Cypress e2e specs for these apps, or fixing bugs and jank in these apps — even if the user just says "the admin UI" or "the teacher dashboard". The stack is frozen on Ext 6.2.0.981: this skill encodes the house patterns to converge on and the upgrades/migrations to explicitly NOT attempt.

vast-ai Skill

Take a new Vast.ai base-image (derivative / pytorch-nested / external) from nothing to a live-GPU-tested working image + usable launch template. Use when adding a new image to this repo. The human picks the class; the agent scaffolds with the imagegen generator, fills the fenced residue to a clean lint, then builds, live-GPU tests, and iterates on real failures (the qa-fix loop, human approving each fix) until the image passes.

johnhuang316 Bundle

Use when code-index-mcp implementation is complete and a version bump, release notes, tag, package publication, or GitHub release is being prepared.

johnhuang316 Bundle

Use when validating code-index-mcp language strategies, tree-sitter symbol extraction, called_by or import baselines, or sample projects under test/sample-projects after parser and indexing changes.

sunpeak-ai Skill

Use when testing MCP servers -- e2e tests with the sunpeak inspector, visual regression testing, live testing against real ChatGPT, multi-model evals, Playwright configuration, or scaffolding test infrastructure with "sunpeak test init". Works with any MCP server (Python, Go, TypeScript, etc.), not just sunpeak projects.

emdashcodes Bundle

Guide for creating effective Agent Skills in Claude Code. This skill should be used when users want to create a new Skill (or update an existing Skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations. Use when working specifically with Claude Code's Skills system.

shren207 Bundle

LLM 추상화 계층, 프로바이더 어댑터, 가격표, 예산 가드 등 LLM 관련 작업이면 무조건 이 스킬. 모델 변경, 프로바이더 추가, 비용 계산, 토큰 카운트 등 모든 LLM 인프라를 다룬다. Triggers: "LLM 모델 변경", "프로바이더 추가", "비용 추정", "예산 가드", "pricing table", "모델 비교", "LLM 비용", "토큰 사용량", "모델 추가", "LLM 설정", "Gemini", "OpenAI", "API key", "model pricing", "budget cap", "token count", "adapter", "factory".

shren207 Bundle

카드 분할 로직, LLM 기반 Split, nid 승계 전략, 텍스트 파서, 비용 추정, 배치 분할 등 카드 분할의 모든 측면을 다룬다. Triggers: "Split 결과가 이상해", "파서 버그", "분할 미리보기", "nid 승계", "Cloze 번호 리셋", "컨테이너 파서", "카드 분할", "atomic card", "분할 비용", "split cost", "배치 분할", "batch split", "분할 적용", "split apply", "분할 검증", "스타일 보존". Covers Split logic, nid inheritance strategy, text parsers, cost estimation, batch split, and split validation.

shren207 Skill

프롬프트 원격 업데이트 워크플로. 코드 수정 없이 API 호출만으로 AnkiConnect config의 시스템 프롬프트를 업데이트한다. prompts.ts는 폴백일 뿐, 실제 프롬프트는 원격 config가 SoT. Triggers: "프롬프트 업데이트", "원격 프롬프트", "프롬프트 동기화", "프롬프트 푸시", "시스템 프롬프트 변경", "시스템 프롬프트 바꿔", "syncing prompts", "prompt update", "prompt sync", "push prompt", "update system prompt", "remote prompt update".

shren207 Bundle

프롬프트 버전 관리, A/B 테스트, 이진 패턴 힌트, 품질 추적 등 프롬프트 시스템 전반을 다루는 스킬. 프롬프트 관련 질문이면 무조건 이 스킬을 먼저 확인할 것. 원격 프롬프트 업데이트 절차는 syncing-prompts 스킬 참조. Triggers: "프롬프트 버전 관리", "A/B 테스트 만들어", "SuperMemo 규칙", "Cloze Enhancer", "프롬프트 성능", "카드 길이 기준", "이진 패턴", "실패 패턴 분석", "시스템 프롬프트 구조", "system prompt 타입", "프롬프트 마이그레이션", "히스토리", "반려 사유", "활성 버전", "메트릭".

bentsolheim Skill

General software architecture principles — extending the existing domain model instead of creating parallel per-feature vocabularies, separating generic capability from feature-specific judgment, naming from the ubiquitous language, and ranking model coherence above change isolation. Use when designing, architecting, or reviewing features, APIs, services, or data models — any time new types, endpoints, resources, or concepts are about to be introduced into an existing system.

brajendrak00068 Skill

Agentic AI video editor: send a prompt and an autonomous editor plans, executes, verifies, and exports the edit — viral clips, captions, vertical reframe, chroma key, audio cleanup, B-roll, motion graphics, MP4 export.

subjadeites Bundle

Codeflow streams coding agent sessions (Claude Code, Codex, Gemini CLI, etc.) to Discord or Telegram in real-time. Use when invoking coding agents and wanting transparent, observable dev sessions — no black box. Parses Claude Code's stream-json output into clean formatted messages showing tool calls, file writes, bash commands, and results with zero AI token burn. Use when asked to "stream to Discord", "stream to Telegram", "relay agent output", or "make dev sessions visible".

necmttn Skill

Surplus-quota training loop over the ax graph - the agent burns the remaining 5h/7d plan-quota window on self-improvement: locking pending verdicts, filling briefs, backtesting routing classes, minting proposals, running worktree experiments, and drafting upstream issue reports. Triggers when the user says "/dojo", "enter the dojo", "dojo time", "train overnight", "burn my surplus quota", "dream mode" (legacy name), or invokes /loop /dojo. Requires ax (axctl) on PATH. ax uses embedded DuckDB; no database daemon is required. Do NOT auto-trigger on unrelated work or when the user merely mentions quotas.

ali5ter Skill

Performs a deep review of the Claude Code plugin, skill, or sub-agent defined in the current project against official best practices. Documents findings as GitHub issues and writes a prioritised fix plan to the project CLAUDE.md. Use when the user says audit this plugin, review this skill, check this agent, or audit addon.

shren207 Bundle

USE THIS SKILL for anything related to embeddings, vector similarity, or semantic search. Triggers: "임베딩 생성", "코사인 유사도", "캐시 어디에", "임베딩 상태", "의미 유사도", "text-embedding-3-large", "openai embedding", "벡터 차원", "3072", "임베딩 캐시", "embedding 마이그레이션", "레거시 캐시", "embedding fallback", "유사도 검사 임베딩 모드", "임베딩 배치", "rate limit 429". Covers OpenAI embedding API, file-based cache with migration, and text preprocessing.

hmakt99 Skill

Trust metadata for files, memories, and skills — check before you trust, stamp what you verify. A stamp costs ~15 tokens; re-verifying costs 15,000.

volcengine Skill

Create or update a pull request following verl project conventions.

volcengine Skill

Create or update a GitHub issue following verl project conventions.

volcengine Skill

Upgrade, downgrade, or re-pin a Python dependency (vllm, sglang, torch, transformers, flash-attn, etc.) in verl's universal uv.lock flow.

hamelsmu Bundle

Work against a live local Jupyter notebook kernel. Use this when an agent needs a Jupyter-like in-memory REPL, wants to inspect or edit a notebook while keeping the kernel alive, or needs an explicit verification pass at the end.

git-ai-project Skill

Analyze AI prompting patterns and acceptance rates

wolfe-jam-faf-skills Skill

Master the IANA-registered .faf format (application/vnd.faf+yaml) — the typed, portable context that makes any AI understand your project. Covers scoring internals, MCP server config, bi-directional sync, and the full 21-slot model. The mechanic's manual for expert-level control. New to FAF? Start with faf-context to hit 100%, then come here to go deep. See also faf-wizard for the done-for-you path.

skyvern-ai Bundle

Automate any website with AI-powered browser automation. Use when the user needs to interact with a website like filling forms, extracting data, downloading files, logging in, or running multi-step workflows. Skyvern navigates sites it has never seen before using LLMs and computer vision. Integrates via Python SDK, TypeScript SDK, REST API, MCP server, or CLI.

necmttn Skill

Guided experiment-loop retrospective over the ax agent-experience graph. Walks the user through their open proposals (accept-with-scaffold or reject), pending verdicts (confirm the suggested verdict or override), and recent harness-hook effectiveness signal. Triggers when the user says "let's do an ax retro", "ax retrospective", "review my ax proposals", "triage proposals", "experiment loop status", "lock pending verdicts", "hook effectiveness review", "intervention review", "self-improvement session", or invokes /ax:retro. Reads/writes via the local `ax improve` and `ax hooks` CLIs. Do NOT auto-trigger on unrelated work.

necmttn Skill

Install + verify ax (the agent experience layer). Triggers when the user says "install ax", "set up ax", "ax not found", "ax doctor", "is ax running", "fix ax install", "first-time ax setup", or any setup question about the ax CLI / skills / daemon. Walks the install via the install.sh + skills.sh + first ingest, validates with `ax doctor`, and points the user at ax:retro (experiment loop) and ax:extract-workflow (reconstruct workflow behind shipped artifacts).

necmttn Skill

Write the agent-generated narration of the current session - the reviewable story of what changed, including what never reaches a PR (user corrections, abandoned attempts, tool failures). Triggers on "narrate this session", "summarize what changed", "write the session story", "narrate what we did", "session narration". Output is .ax/narrations/<session-id>.json for the ax studio narration view. Do NOT fire on "summarize this file" or generic recap questions answered inline - this skill writes a structured artifact.

necmttn Skill

Deep retro of retros - investigation pass that surfaces improvements across older retros and the current ax setup. External AI agent drives the reasoning with high thinking enabled. Triggers when user says "deep retro", "investigate ax", "what should I fix in my setup", "retro of retros", or invokes /ax:retro-meta. Use after `ax:retro` if proposals are sparse or you want broader exploration.

necmttn Skill

Definition-of-done checklist for shipping a new write, signal, table, edge, or query to the ax graph. Use when adding a SurrealDB table/edge/field, an ingest derive-stage, a new analytic query, or any new ax capability - before opening the PR. Ensures every write gets an on-demand read AND a proactive (agent-facing) read AND docs/distribution, not just the write. Triggers on "ship", "new signal", "new table/edge", "new lens/query", "wire this up", "is this done", or finishing an ax feature branch.