AI & ML

stackhawk Skill

Guides authoring and maintaining agent skills in this repository — enforcing Anthropic best practices, bumping versions, regenerating Cursor rules, and keeping the plugin structure consistent. Use when editing SKILL.md files, reference files, or plugin manifests; creating a new plugin; or reviewing a skill for compliance with authoring standards. Do NOT trigger for general coding tasks, security scanning, or platform API queries.

tuya Bundle

Agent-friendly non-blocking serial log capture for TuyaOpen devices. Runs tos.py monitor in the background, writes logs to file, and lets the agent tail or stop it at any time. Use when the agent needs to capture device logs without blocking, run a debug session, or monitor UART output during a build-flash-analyze loop. 后台串口日志、非阻塞监控、设备日志捕获、agent日志分析。

umbraco Skill

This repo's branch model, and the only holder of it. Merges a PR with whichever strategy the model calls for, opens a PR onto the correct base for its line, and starts a work branch rooted correctly. The integration branches, the release base and the merge strategy are PRIVATE — a caller asks for an outcome and never learns a branch name or a strategy. Command-only: it answers no reads. Called by name with (action, context-json) by a service, never by a loop. NOT for direct use — never select it from a description match.

umbraco Skill

An isolated place to build and test one change in Umbraco.Automate — a git worktree under `.claude/worktrees/`, rooted on the target line's integration branch, with the local config files a build needs — and the teardown that removes it. Prepared and torn down by `ops-change`, never by a loop. Called by name with (action, context-json). Not model-invoked.

vgtitov Bundle

Разработка и код-ревью на 1С:Предприятие 8.3 (BSL) с проверкой по РЕАЛЬНОМУ коду через MCP, а не по памяти. Используй всякий раз, когда пишешь, правишь или ревьюишь BSL-код, работаешь с общими модулями, формами, запросами, СКД, регистрами, ролями; реализуешь задачу по ЧТЗ/тех-проекту; ищешь, как устроена типовая конфигурация или что уже есть в расширении; или отвечаешь на технический вопрос «как сделать в 1С» и «где это в коде». Срабатывай даже без слов «BSL/код/ревью», если речь о реализации/правке в 1С. Нужны ДАННЫЕ живой/файловой базы (посчитать записи, прочитать регистр, подключиться и проверить через OData) — это НЕ задача этого скилла: вызывай `1c-tester` — он выбирает уровень в фиксированном порядке: чувствительные данные (персональные, платёжные, зарплатные) предпочтительно через ось доступа даже при разовом чтении, а без публикации допустим и раннер, но тогда результат обезличивает человек РУКАМИ перед отчётом, показом вовне или передачей AI (§B.2 `references/data-access-verification.md`); разовое чт

vgtitov Bundle

Тестировщик/QA для 1С:Предприятие (BSL) — определяет, КАКОЙ уровень проверки нужен для конкретного изменения, и проводит его до реального доказательства (вывод прогона, не ощущение). Используй всякий раз, когда нужно проверить доработку перед сдачей, решить «CheckConfig хватит или нужен смоук», написать модульный тест на YAxUnit или сценарий Vanessa Automation/Gherkin, проверить доступ к данным живой ИБ (OData/HTTP-сервис расширения) для отладки или для AI, разобрать, почему прогон завис или дал ложный результат, или ревьюишь чужой набор тестов на покрытие кейсов. Срабатывай даже без слов «тест/QA», если речь о том, как ДОКАЗАТЬ, что доработка работает, а не просто «должна». Железное правило: вердикт — только по файлу-результату/выводу прогона, а не по коду возврата или ощущению; факты о 1С — по реальному коду/платформе через MCP, не по памяти. Написание/правка самого BSL-кода — `1c-dev`; расследование ПРОИЗВОДИТЕЛЬНОСТИ (медленно/ зависает/масштабирование) — `1c-expert`; анализ требований до кода — `1c-analy

vgtitov Bundle

Анализ и архитектура задач 1С: разбор задачи/ЧТЗ, какие контуры/базы существуют и что с чем связано, что затронет изменение, где уже реализован функционал; подготовка качественного ЧТЗ (ожидаемое поведение + модель данных + критерии приёмки) и архитектурных решений. ОБЯЗАТЕЛЬНО используй, когда анализируешь или уточняешь задачу по 1С, определяешь контур и базу, оцениваешь влияние доработки, готовишь/проверяешь ЧТЗ, формулируешь уточняющие вопросы, решаешь нужен ли архитектор, или проектируешь на уровне архитектуры (не написания BSL). Срабатывай даже без слов «анализ/архитектура», если речь о понимании задачи, контуров, влияния или подготовке требований. Главное правило: искать в ОБОИХ слоях контура (конфигурация + расширение) и проверять по РЕАЛЬНОМУ коду через MCP, не по памяти. Написание/ревью BSL — скилл `1c-dev`.

pedrohcgs Skill

Draft a structured preregistration document (OSF, AsPredicted, or AEA RCT Registry style) from a research spec or free-form study description. Output is a Markdown file with hypotheses, design, sampling plan, analysis plan, exclusions, and inference criteria — annotated with MUST / SHOULD / MAY clarity flags. Use when user says "preregister", "draft a preregistration", "OSF preregistration", "AsPredicted", "AEA RCT registry", "PAP", "preanalysis plan", or before launching an experiment / data collection / analysis on data the analyst has not yet seen. NOT a registry submission tool — produces a document the user uploads to OSF / AsPredicted / AEA themselves.

pedrohcgs Skill

Run Chain-of-Verification (CoVe) on a draft or a block of text with factual claims. Spawns the `claim-verifier` agent in a forked (fresh) context so it never sees the draft — then reports which claims are supported, contradicted, or unverifiable. Use when user says "verify these citations", "check the claims in X", "did I hallucinate anything", "fact-check this draft", "run CoVe on this", or after any text generation that asserts facts about papers, datasets, or numerical results. NOT for style/grammar review (use `/proofread`) or substance review (use `/review-paper`).

zai-org Bundle

Generate captions (descriptions) for images, videos, and documents using ZhiPu GLM-V multimodal model series. Use this skill whenever the user wants to describe, caption, summarize, or interpret the content of images, videos, or files. Supports single/multiple inputs, URLs, local paths, and base64 (images only).

zai-org Bundle

A skill that uses GLM-V native grounding capabilities for coordinate conversion, bounding-box visualization, and more. GLM-V native grounding can locate any target specified by the prompt in an image and output relative coordinates normalized to 0-1000 based on image size. Coordinate formats include 2D bounding box (default), 2D points, and 3D bounding box. GLM-V also supports spatiotemporal localization and tracking of multiple prompt-specified targets in videos, outputting 2D bounding boxes per second.

zai-org Bundle

Analyze images/videos and generate professional prompts for text-to-image and text-to-video AI tools (Midjourney, Stable Diffusion, DALL-E, Sora, Runway, Kling, Pika). Use when the user wants to generate prompts from reference images/videos, create AI art prompts, or get prompt engineering suggestions from visual content.

zai-org Bundle

Screen and evaluate resumes against criteria using ZhiPu GLM-V multimodal model. Reads multiple resume files (PDF/DOCX/TXT), compares against user-defined screening criteria, and outputs a Markdown table with pass/fail analysis. Use when the user wants to filter resumes, compare candidates, or batch-evaluate job applications.

zai-org Bundle

Write a textual content based on given document(s) and requirements, using ZhiPu GLM-V multimodal model. Read and comprehend one or multiple documents (PDF/DOCX), write a content in Markdown format according to the specified requirements. Use when the user wants to draft a paper/article/essay/report/review/post/brief/proposal/plan, etc.

builderio Bundle

Use when adding or following the red/yellow/green final status block convention for agent responses, especially by installing managed AGENTS.md or CLAUDE.md instructions.

builderio Bundle

Use when asked to watch, babysit, audit, review, compare, or fix another agent's work from a Codex session ID, Claude Code session/transcript, chat/thread link, PR, branch, log, or pasted run summary. Monitor until the other agent is done or blocked, reconstruct what the user asked, inspect what the agent actually changed and verified, report gaps, and optionally make scoped fixes when the user authorizes repair.

builderio Bundle

Use when long-running or parallel agent work must respect 5-hour and weekly usage limits by checking usage between waves, pausing near the cap, and resuming only when the window is clear.

automateip Bundle

Agent-skills version of a colliding skill name — this one must win.

automateip Skill

A skill that only exists in the agent-skills source.

manifold-works Skill

Generates and applies EF Core migrations after domain or DbContext model changes in a dotnet-vite-webapp-template app, with dev vs production apply rules. Use when the user invokes /add-migration, changed entities/DbContext, or needs dotnet ef migrations add/update guidance.

perry-lynn Bundle

Codex 可配置双对话开发协作。用户要求两个对话、双模型开发或模型路由时，按用户指定的 Planner 与 Executor、模型和推理强度互读交接、验证与重新规划；未指定时默认 Astra low 规划、Luna high 实施。

spencer1o1 Skill

Makes the current agent the project boss: owns planning, vault Decisions, docs/BRIEF.md, and delegation. Use when the user says you are the boss, @boss, or wants Cursor to run the phase.

spencer1o1 Bundle

Installs Cursor stop/afterFileEdit verify hooks (format, lint, typecheck) from this skill's scripts/ template. Use when the user asks to add verify hooks or auto-verify on agent stop.

timurugulava Bundle

Создаёт мемы-картинки под идею/тезис пользователя: подбирает подходящий популярный шаблон (Drake, Distracted Boyfriend, This is Fine, Galaxy Brain, Spongebob Imagination, Two Buttons, DiCaprio Cheers и ещё 20+), формулирует конкретные русские надписи, рисует через Nano Banana 2 (Google Gemini) по MCP-серверу `nanobanana`. Поддерживает опциональный авторский голос-профиль и нейтральный режим по умолчанию. Use when пользователь пишет: «сделай мем», «мем про X», «мем для канала», «мем на тему», «нарисуй мем», «придумай мем», «хочу мем», «мем-картинка», «meme-maker», «/meme-maker». Do NOT use for обычных картинок без мем-формата, рекламных креативов с ЦА и УТП, презентаций из слайдов, PDF-гайдов, обложек постов без мем-шаблона.

tscarpe Bundle

AI 测试智能体编排：用例生成 → 静态代码一致性检测 → test 环境浏览器功能测试 → 报告生成。子命令（cases/static/exec/spec/report）独立可重跑，两道人工关卡（用例审核、报告复验）；通过用例可资产化为 Playwright spec，回归轮跑 runner 零 agent token。Use when 用户要求 AI 测试、生成测试用例、执行测试、提测验证、回归测试、生成回归脚本、spec 资产化，或说 AI testing、test generation、browser E2E、playwright spec。用法：/sdlc-test cases|static|exec|spec|report <需求名>

apizanti-uxr Skill

End-of-day log. Requires all five MCPs healthy first (mcp-healthcheck). Gathers today's calendar, Drive transcripts, Jira, and Slack, scores items, writes to the local wiki created by daily-rituals, and asks once about Jira actions. Use when called by daily-rituals /eod.

apizanti-uxr Skill

Morning startup and midday check. Handles /startup, /morning, and /midday. Checks MCP health, pulls calendar, last EOD focus, open Jira, and carry-forwards, then renders a Cursor Canvas. Use when called by daily-rituals /startup, /morning, or /midday.

apizanti-uxr Skill

Checks that all critical MCP servers are live before daily rituals. Required before /startup, /morning, /midday, /eod, and /brief. Tests Slack, Atlassian, Google Workspace, Dataverse, and Smartsheet. Attempts recovery, then stops if any source is still down.

aresiddharthareddy Skill

Keeps agent replies and tool use lean. Use for all user-facing responses and multi-step tasks. Reduces chat tokens without hiding critical information.

aresiddharthareddy Skill

Step-by-step guardrails for weaker or fast models. Use on complex tasks, multi-file changes, or when quality slips. Forces verify-before-write and single-step focus so smaller models match larger ones on correctness.

asunalabs Skill

Router for the waves work framework. Use when starting work in a waves workspace/repo, when the user types /waves with no specific sub-skill, or asks 'how do we work here'. Routes to waves-init (workspace setup), waves-plan (task breakdown), waves-run (parallel execution via agent teams or subagents), waves-status (coordinator blurb), platform-guide (third-party dashboard how-tos).

asunalabs Skill

Executes a master plan wave-by-wave: fans the largest unblocked task set out across parallel workers — an agent team (shared task list, teammate messaging) when enabled and the wave benefits, subagents otherwise — with worktree isolation and explicit file ownership, then merges, verifies live, updates the plan, and emits a coordinator blurb at the wave boundary. Use when running a plan, launching the next wave, orchestrating parallel agents, or spawning teammates for plan execution.

asunalabs Skill

Initialize a workspace or repo for AI-agent work: scan repos, index their rules, run the mandatory scoping interview, generate workspace CLAUDE.md + plans/00-master.md. Use for a new task/workspace, when a repo has no agent instructions, or to refresh a stale workspace setup.

colindmurray Skill

Use when delegating a coding task to a CLI coding agent (Claude Code, Codex, Gemini, OpenCode, Pi) — investigations, multi-step refactors, parallel issue work, PR reviews. Direct mode for quick (<5 min) tasks; pair with long-running-task for background execution >5 min.

cyanheads Skill

Exercise mcp-reporter's collected reports through in-memory, stdio, HTTP and compiled Node entry points.

dangxingyu Bundle

Formalize a mathematical or theoretical research problem, write a self-contained cyclic research prompt, launch independent construction and stress-test subagents, and iterate through proof-gate extraction, independent auditing, and synthesis. Use when the user asks to launch theory agents, explore a theorem or separation with subagents, turn a research question into a multi-agent proof search, apply a cyclic prompt to theoretical work, or rigorously seek a proof, counterexample, lower bound, construction, or impossibility result.