AI & ML

curiositech Bundle

Track real-time API cost accrual during LLM execution. Activate on 'cost tracking', 'token usage', 'API costs', 'budget monitoring', 'usage metrics'. NOT for cost estimation, pricing tiers, or billing systems.

vapiai Bundle

Create, improve, or audit Vapi voice agent and Squad system prompts for production phone and web based voice agents. Use when the user wants help designing a Vapi assistant prompt, multi-assistant Squad prompt set, refining an existing prompt, creating prompt sections, building an intake or handoff workflow, improving tool-use instructions, adding guardrails, or optimizing voice-agent behavior for brevity, turn-taking, error handling, caller data collection, escalation, handoffs, and spoken formatting.

motherduckdb Bundle

Discover MotherDuck databases, tables, columns, shares, and sample data to understand an available dataset.

motherduckdb Bundle

Design or implement MotherDuck analytical schemas and transformation models, including grain, types, and materialization.

microsoft Skill

Use when the user wants the dsh agent to self-improve from past usage, asks about a nightly/offline 'sleep' or 'dream' cycle, skill/memory consolidation, or says things like 'make my agent better the more I use it', 'review my past sessions', 'learn my preferences', 'consolidate what you learned', 'run the sleep cycle', or wants to schedule background self-optimization. Drives the skillopt_sleep engine through the skillopt_* tools: harvest past sessions -> mine recurring tasks -> replay via a selected backend -> consolidate validated skills behind a held-out gate.

robbyt Skill

Get Codex's code review of git changes via the Codex MCP server. Trigger when user wants a second opinion on code changes ("have Codex review my changes", "get code review from Codex", "review this diff with Codex"), or as a final check before committing.

robbyt Skill

Get Codex's review of an implementation plan before the user starts building — trigger when they want Codex (or a named GPT model like gpt-5.6) to review, critique, or pressure-test a plan. Applies to any plan-shaped artifact — a plan file, plan-mode plan, a migration/rearchitecture/integration/checkout write-up, or a doc describing how they intend to do something (pasted, at a repo path, or just described). Fire on any second-opinion phrasing — "poke holes in it", "sanity-check my plan", "run this past Codex", "get Codex's take", "flag the biggest risks", "look this over before I start", "did I miss edge cases, testing, or a rollback path?". The point is catching gaps, risks, missing steps, and better alternatives ahead of implementation. Do NOT use for reviewing already-written code or diffs (that's diff-review), mapping an existing codebase's architecture (codebase-analysis), or web/research questions.

robbyt Skill

Codebase analysis via the Codex MCP server with a read-only sandbox. Trigger when user needs architecture overview ("analyze this codebase with Codex", "have Codex map dependencies"), onboarding to unfamiliar code, understanding legacy systems, or identifying technical debt.

curiositech Bundle

Validates that a DAG node's output matches its declared JSON schema before passing to downstream nodes. The glue that makes multi-agent DAGs reliable. Use when checking output contract compliance, generating validation schemas from descriptions, or debugging contract mismatches between nodes. Activate on "validate output", "output contract", "schema validation", "contract mismatch", "output doesn't match". NOT for content quality assessment (use dag-quality), skill grading (use skill-grader), or general JSON schema work outside DAG context.

curiositech Bundle

Expert photography composition critic grounded in graduate-level visual aesthetics education, computational aesthetics research (AVA, NIMA, LAION-Aesthetics, VisualQuality-R1), and professional image analysis with custom tooling. Use for image quality assessment, composition analysis, aesthetic scoring, photo critique. Activate on "photo critique", "composition analysis", "image aesthetics", "NIMA", "AVA dataset", "visual quality". NOT for photo editing/retouching (use native-app-designer), generating images (use Stability AI directly), or basic image processing (use clip-aware-embeddings).

rsmdt Skill

Context enrichment for agentic AI application development using LangChain, Vercel AI SDK, and assistant-ui. Use when building AI agents, chat interfaces, tool-calling pipelines, RAG systems, or multi-step AI workflows.

creekmoon Bundle

为代码仓库生成或重构根目录 README。只要用户提到 README、项目介绍、项目文档、架构说明、新人接手、仓库导览、模块边界、模块协作，或工具/CLI/MCP/SDK 的安装、启动、用法、开箱即用、配置说明，就应该优先使用本技能。先判读者再选内容：读者要用起来就用法优先，读者要接手就边界与流程优先，两者都要就混合组织；章节与详略从项目实际长出来，不套固定骨架。不要写成业务说明书、项目知识图谱、项目记忆、完整 API 文档、代码说明书或深度研究报告。

vishalsachdev Bundle

Convene a 3-model council (Claude + GPT via codex CLI + Gemini CLI) on a high-stakes decision. Forces cross-critique between members and surfaces where they actually disagree, breaking Claude's default agreeableness. Use when the user asks to "convene a council", "get a second opinion", "ask GPT and Gemini", "what would other models say", or has an architecture / strategy / hiring / pricing decision where being wrong is expensive. Skip for factual questions, code with one right answer, or anything premortem-shaped (route to premortem skill instead).

fan648512 Bundle

当用户的提示词（prompt）过于简短、模糊或缺少要素，需要扩写、改写为结构清晰、目标明确、可直接复制执行的高质量提示词时使用；也用于优化已有的系统提示词或人设。触发词包括"帮我优化这个提示词""增强提示词""这句提示词怎么写更好""扩写一下""润色提示词""这个 prompt 怎么写""review 这个 prompt"，以及"为什么 AI 输出不符合预期、该怎么改指令"。不适用于：纯知识提问、贴报错求调试、闲聊、要求逐字保留的法律或合规文本。

tibsfox Bundle

Interrupt controller converting polled to proactive agent execution. Per-runtime strategies, configurable thresholds, Deacon heartbeat supervision. Fights LLM passivity bias.

rewse Bundle

Create, improve, and review Kiro steering files based on prompt engineering best practices. Use when the user wants to create a new steering file, improve or optimize an existing one, review steering files for quality, audit their steering setup, or asks about steering file best practices. Also trigger when the user mentions "steering", "steering file", ".kiro/steering", "project rules", "coding standards file", "conventions file", or wants to establish persistent AI guidelines for their project.

rewse Skill

Personal browser automation preferences and configuration. Use when the user needs web automation, browser interaction, headed mode, or connecting to the user's Google Chrome via CDP. Also use when agent-browser or playwright-cli is about to be used, to ensure correct engine selection and connection settings.

yixuan-13 Bundle

Headless control of a locally-running ComfyUI server (default http://127.0.0.1:8188) via REST API. Model-agnostic — runs ANY workflow JSON, edits ANY node's parameters, lists installed nodes/models, uploads inputs, polls job status, downloads outputs. Use when the user wants to generate images/video/audio with ComfyUI, manage models, or batch-generate. Designed as a pure-file fallback for the artokun/comfyui-mcp MCP server (if MCP fails, this skill still works with just curl/python).

tsaru23 Skill

トークン消費を節約するためのモデルオーケストレーション。複数のIssue・タスク・機能をまとめて 実装する依頼、「タスクを振り分けて」「安いモデルで済ませて」「オーケストレーションして」 「Issueを消化して」「コスパよく実装して」といった依頼では必ずこのスキルを使うこと。 最上位モデル(Fable)はアドバイザーとして設計・担当割り当て・レビューに専念し、実装は 難易度に応じて Opus / Sonnet / Haiku / 人間 に割り当てる。品質不足の成果物は担当を 上位モデルに切り替えて再実行する。単発の小さな修正1件だけの依頼には使わない。

ax128 Bundle

AegisGate Agent Skill

dtsola Bundle

OpenClaw context check / context optimization (Context Budget). Core goal = context optimization: reads the models currently enabled in this installation, reads each model's published context window from its vendor's official source, and proposes a window = 60% of the vendor spec. It writes nothing until the user confirms; then it patches only the window field and reports the result. This skill never runs automatically and creates no scheduled jobs. Use ONLY when the user explicitly asks, with one of these exact intents: 上下文检查 / 上下文优化 / 检查一下模型上下文 / 把上下文窗口配一下 / 设置上下文窗口, or in English: context window check, configure context window, context optimization. Do NOT activate on generic talk about context, memory, prompts, or token usage. 中文：按「厂商标称窗口 × 60%」设置已启用模型的上下文窗口；流程 = 检测（只读）→ 决策（用户回一个数字）→ 执行（确认后写入）。 不做：maxTokens 等其它参数、压缩阈值（保持系统默认）、未在用模型的默认配置、自动/定时运行、修改历史与审计留痕。

easyindie Skill

软件实现角色。负责功能开发、缺陷修复、重构落地和必要测试，目标是交付可运行、可验证、可维护的代码。

easyindie Skill

Devops Agent

easyindie Skill

技术调研角色。负责收集外部资料、验证事实、对比方案、补充上下文，为架构和实现提供可信输入。

easyindie Skill

软件架构角色。负责需求抽象、技术选型、模块划分、系统边界、风险识别和实现路线，不负责直接拍代码。

unsafe9 Bundle

Transfer a self-contained task to a separate, persistent agent surface that keeps working on its own — Codex app (via a `codex://` deeplink), Claude Code in tmux, or Codex CLI in tmux. Composes a self-contained brief, launches the surface, and reports where it is running; control moves to that surface and the work continues independently in its own app or terminal window. The point is to set up the work to live on elsewhere, not to wait on it here. Use when the user wants to continue work in another app or terminal, open or launch a task in Codex, spin up an autonomous background or tmux agent, or says 'handoff', 'continue this in tmux', 'open it in the Codex app', 'run it in the background'. When the user says codex, default to Codex app; use Codex CLI only when explicitly named.

unsafe9 Bundle

Write, edit, or review durable agent instructions in CLAUDE.md, AGENTS.md, skills, agent prompts, or context-injecting hooks. Also use to analyze agent mistakes and propose prevention.

easyindie Bundle

Use acpx as a headless ACP CLI for agent-to-agent communication, including prompt/exec/sessions workflows, session scoping, queueing, permissions, and output formats.

easyindie Skill

质量保障角色。负责测试设计、回归验证、缺陷归类、发布建议，目标是让团队知道“哪里有风险、能不能发”。

easyindie Skill

产品体验角色。负责用户流程、信息架构、交互反馈和可用性优化，确保功能不仅能做，还能被顺畅使用。

easyindie Skill

研发总指挥。需求拆解、子Agent调度、进度推送、质量门禁、交付整合

dasdigitalemomentum Skill

Run Playwright-MCP browser acceptance, agent-observed walkthroughs, or user-attended walkthroughs through the role routing appropriate to each mode, returning a compact evidence-referenced digest.

dasdigitalemomentum Bundle

Record environment, harness, or tooling issues that block or degrade agent work into an append-only, deduplicated log so the environment can be improved incrementally.

viscous106 Skill

Use when adding a new skill, agent, or slash command to the claude-skills repo — places the file correctly and gets the frontmatter right.

yufanyang214-lab Bundle

Archive or recall durable facts, procedures, decisions, and events across Codex CLI sessions using a standalone evidence/object/bright-dark-ledger workspace rather than Codex memories.

beyond-entity Bundle

Use Beyond Entity as persistent AI architecture memory to recover project context, review or evolve architecture, implement code from current BE designs, and synchronize architecture after code changes through MCP. Use for Beyond Entity project work, not unrelated database design.