mcpbeat Sign in

Claude Skills

The open format is called Agent Skills and works in Claude Code, Codex, Cursor and other agents — most people know it as Claude Skills.

Every Agent Skill we could find on GitHub, deduplicated by content. 86 961 files from 2 125 authors, of which 68 678 are unique — the rest is the same skill repackaged into someone else's repository. For each one: what it weighs in tokens, whether it ships runnable scripts, and which MCP servers it needs.

68 678
unique skills
out of 86 961 files found on GitHub
18 283
are copies
same content, someone else's repository
1 749
tokens, median
what a typical skill costs you in context
8 467
name collisions
two skills with one name cannot sit side by side

13 801–13 860 of 68 678

page 231 of 1 145
Udf Optimize Cudf
by NVIDIA

Iteratively optimizes a cuDF RapidsUDF implementation for GPU performance. Use after testing and benchmarking with udf-benchmark. Runs a loop of profiling, optimizing, testing, and benchmarking until performance converges or the iteration budget is exhausted.

1.8k tokens
Udf Convert To SQL
by NVIDIA

Assists with converting an Apache Spark UDF to a functionally equivalent Spark SQL expression. This is step 2 of 3 in the UDF conversion workflow (udf-gen-test -> udf-convert-to-sql -> udf-benchmark). Use this skill when you have a CPU UDF with a unit test and need to convert it to SQL for GPU acceleration.

1.1k tokens
Udf Judge Conversion
by NVIDIA

Reviews generated UDF tests and GPU/SQL implementations for robustness, anti-cheating, and GPU execution integrity. Use when the user requests a judge/review-agent pass, or when manually reviewing a completed conversion.

1.3k tokens
Udf Convert To Cuda
by NVIDIA

Assists with converting a non-aggregating Apache Spark UDF to a native CUDA RapidsUDF using JNI and libcudf. This is step 2 of 3 in the UDF conversion workflow (udf-gen-test -> udf-convert-to-cuda -> udf-benchmark). Use this skill when you have a CPU UDF with a unit test and need to convert it to a native CUDA implementation. Prefer udf-convert-to-cudf unless a CUDA implementation is necessary for performance or correctness, or if requested by the user.

2.2k tokens scripts
Udf Benchmark
by NVIDIA
vendor

Assists with benchmarking and profiling the performance of an Apache Spark UDF on the GPU. This is step 3 of 3 in the UDF conversion workflow (udf-gen-test -> udf-convert-to-* -> udf-benchmark). Use this skill when you have a CPU UDF and a RapidsUDF or SQL implementation, and need to benchmark the performance of the CPU UDF against the GPU implementation.

755 tokens
Udf Gen Test
by NVIDIA

Assists with generating a unit test for an Apache Spark UDF. This is step 1 of 3 in the UDF conversion workflow (udf-gen-test -> udf-convert-to-* -> udf-benchmark). Use this skill when you have a CPU UDF and need to create a unit test for the UDF before converting it into a GPU-compatible implementation.

1.7k tokens scripts
Cali
by callstackincubator

Use when working in the Cali repository or when you need to run, extend, or debug the Cali CLI for mobile React Native and Expo workflows. Covers Cali commands (`qa`, `review`, `perf-review`, `dev`), the shared `cali-context.json` contract, local mode selection, required local CLIs, provider setup, and CI integration patterns.

731 tokens
Fboss Review
by facebook

Comprehensive FBOSS code review with 11 parallel reviewers (5 generic + 6 FBOSS-specific). Covers reliability, engineering, code quality, silent failures, agent architecture, SAI/SDK, FSDB/thrift_cow, platform/config, testing, and cross-cutting design. Self-contained alternative to /review-diff for FBOSS diffs. Findings shown to user only, never auto-posted.

1.1k tokens
Fboss Code Standards
by facebook

FBOSS coding standards and patterns. Auto-loaded when writing code in fboss/ to catch architecture violations, SAI/SDK misuse, thrift_cow pitfalls, platform config errors, and testing gaps. For explicit multi-reviewer review, use /fboss-review instead.

1.6k tokens
Debug Agent Hw Test
by facebook

Debug FBOSS AgentHwTest failures - build, run, analyze logs, crash debug, vendor diag shell, hypothesis-driven debugging. Use when running sai_agent_hw_test or multi_switch_agent_hw_test and investigating test failures.

3.1k tokens scripts
Kineto Release
by pytorch
vendor

Update the third_party/kineto submodule in PyTorch to the latest commit from this kineto repo and commit the change. Use when updating the kineto submodule hash for a release.

574 tokens
Qdk Programming
by microsoft
vendor

QDK programming guide for Q#, OpenQASM, and the qsharp/qdk Python API. Use when: writing or editing Q# or OpenQASM code, running quantum programs, resource estimation, circuit generation, noisy simulation, Azure Quantum submission, Qiskit/Cirq/PennyLane interop, or Q# testing and standard library usage.

863 tokens
Kotlin Backend Jpa Entity Mapping
by Kotlin
vendor

> Model Kotlin persistence code correctly for Spring Data JPA and Hibernate. Covers entity design, identity and equality, uniqueness constraints, relationships, fetch plans, and common ORM (Object-Relational Mapping) traps specific to Kotlin. Use when creating or reviewing JPA (Java Persistence API) entities, diagnosing N+1 or LazyInitializationException, placing indexes and uniqueness rules, or preventing Kotlin-specific bugs such as data class entities and broken equals/hashCode.

1.9k tokens
Kotlin Tooling Agp9 Migration
by Kotlin
vendor

> Migrates Kotlin Multiplatform (KMP) projects to Android Gradle Plugin 9.0+. Handles plugin replacement (com.android.kotlin.multiplatform.library), module splitting, DSL migration, and the new default project structure. Use when upgrading AGP, when build fails due to KMP+AGP incompatibility, or when the user mentions AGP 9.0, android multiplatform plugin, KMP migration, or com.android.kotlin.multiplatform.library.

5.4k tokens scripts
Kotlin Tooling Cocoapods Spm Migration
by Kotlin
vendor

Migrate KMP projects from CocoaPods (kotlin("native.cocoapods")) to Swift Package Manager (swiftPMDependencies DSL) — replaces pod() with swiftPackage(), transforms cocoapods.* imports to swiftPMImport.*, and reconfigures the Xcode project.

8.3k tokens
Kotlin Tooling Immutable Collections 0 5 X Migration
by Kotlin
vendor

> Migrate Kotlin (and Java) code from kotlinx.collections.immutable 0.3.x / 0.4.x to the latest 0.5.x. The 0.5.x line renames every copy-returning method on PersistentList / PersistentMap / PersistentSet / PersistentCollection to a participial form per KEEP-0459 (add→adding, removeAt→removingAt, set→replacingAt, put→putting, clear→cleared, …) and version, recompile, and apply the rename each deprecation warning names. Use when the user mentions kotlinx.collections.immutable 0.5.x, PersistentList migration, "Use adding() instead", KEEP-0459, or sees deprecation warnings from kotlinx.collections.immutable.

2.4k tokens
Kotlin Tooling Java To Kotlin
by Kotlin
vendor

> Use when converting Java source files to idiomatic Kotlin, when user mentions "java to kotlin", "j2k", "convert java", "migrate java to kotlin", or when working with .java files that need to become .kt files. Handles framework-aware conversion for Spring, Lombok, Hibernate, Jackson, Micronaut, Quarkus, Dagger/Hilt, RxJava, JUnit, Guice, Retrofit, and Mockito.

1.4k tokens
Kotlin Tooling Native Build Performance
by Kotlin
vendor

> Diagnoses and fixes slow Kotlin/Native compilation and linking in Kotlin Multiplatform projects that target iOS. Use when the user reports slow iOS or shared-framework builds, long linkDebug*/linkRelease* or XCFramework tasks, cold CI builds that re-download the Kotlin/Native toolchain, KSP or other generated code on the native path, transitiveExport usage, or asks for a local-development versus CI build performance plan.

2k tokens scripts
Dd Logs
by DataDog

Log management - search, pipelines, archives, and cost control.

1.3k tokens
Pup
by DataDog

Datadog API CLI with 49 command groups, 300+ subcommands. Skills and domain agents for monitoring, logs, APM, security, and infrastructure.

490 tokens scripts
Dd Apm
by DataDog

APM - traces, services, dependencies, performance analysis.

1.5k tokens
Dd File Issue
by DataDog

File GitHub issues to the right repository (pup CLI or plugin)

2.3k tokens
Dd Monitors
by DataDog

Monitor management - create, update, mute, and alerting best practices.

1.1k tokens
Dd Debugger
by DataDog

Live Debugger - inspect runtime argument/variable values in production by placing log probes on methods. Use when asked what values a function receives, what parameters look like at runtime, or to capture live data from running services without redeploying.

2.7k tokens
Dd Code Generation
by DataDog

Use pup CLI for immediate Datadog operations or generate code for integration into applications

4.3k tokens
Dd Triage Flaky Test
by DataDog

Load when investigating a specific flaky test. Gets history, failure pattern, and category, then recommends fix, quarantine, or escalate.

2.4k tokens
Dd Pup
by DataDog

Datadog CLI (pup). OAuth2 auth with token refresh.

2.2k tokens
Dd Symdb
by DataDog

Symbol Database - search service symbols, find probe-able methods.

781 tokens
Dd Unblock Pr
by DataDog

Load when investigating a failing PR CI pipeline or checking PR health. Attributes each CI failure as flaky, infra, or regression, proposes a targeted action, and reports code coverage.

1.7k tokens
Validate Skills
by microsoft
vendor

Validate that commands documented in skill files actually work. Use when creating, updating, or reviewing skills to ensure all documented commands exit with code 0.

760 tokens
Creating Skills
by microsoft
vendor

Create custom agent capabilities when discovering novel tools, receiving task-agnostic tips from reviewers, or after researching specialized workflows not covered in existing instructions. Teaches structure, YAML optimization for LLM discoverability, and token efficiency.

1.5k tokens
Trx Analysis
by microsoft
vendor

Parse and analyze Visual Studio TRX test result files. Use when asked about slow tests, test durations, test frequency, flaky tests, failure analysis, or test execution patterns from TRX files.

1.8k tokens
Expert Reviewing
by microsoft
vendor

Route code changes to relevant review dimensions based on affected files and folders. Use when reviewing PRs, analyzing code quality, or performing targeted reviews of specific vstest subsystems.

1.7k tokens
Vstest Build Test
by microsoft
vendor

Build, test, and validate changes in the vstest repository. Use when building vstest projects, running unit tests, smoke tests, or acceptance tests, or when deploying locally built vstest.console for manual testing.

1.7k tokens
Source Command Opsx Explore
by Gitruck

Enter explore mode - think through ideas, investigate problems, clarify requirements

1.7k tokens
Source Command Opsx Archive
by Gitruck

Archive a completed change in the experimental workflow

1.3k tokens
Openspec Archive Change
by Gitruck

Archive a completed change in the experimental workflow. Use when the user wants to finalize and archive a change after implementation is complete.

1.3k tokens
Source Command Opsx Apply
by Gitruck

Implement tasks from an OpenSpec change (Experimental)

1.2k tokens
Source Command Opsx Propose
by Gitruck

Propose a new change - create it and generate all artifacts in one step

1.1k tokens
Gtrk Composite
by Gitruck

合成层包装(合成颗粒)——粗剪、B-roll、字幕都定了之后,把选定时间窗做成「合成」:一段时间内的整个画面由一颗代码颗粒接管(画中画拼贴、逐句换画面、对照并排、并列卡片、字幕排版),讲者与 B-roll 作为槽位画进去。三种入口:整片铺满 / 让 AI 判定哪里值得做 / 用户指定区间;候选窗口显示在客户端大纲的 CP 行。当用户想「做合成 / 合成层 / 整片做一下合成包装 / 哪里适合做合成 / 这段做成画中画拼贴 / 11 秒到 32 秒做个合成 / 逐句换画面 / 把字幕排进画面里」时使用本 skill,驱动 `gtrk composite`(plan → review → scaffold → build → lint → lay → bake),别手搓 HTML 铺进工程。

5.4k tokens zh
Gtrk AI Drama
by Gitruck

AI 情景动画分镜稿生成器——把成片 SOP 第③步(B-roll 底轨阶段)分到 AI_DRAMA 车道的 beat,逐个产**四段朴素描述(故事背景 / 角色 / 分镜 / 原文文稿)+ 独立视觉基调段 + 时长预算**,中英双版;Markdown 外壳默认严格兼容 gitruck-ai-drama-desk 解析器,可直接导入工作台,也可拿去任意外部 AI 漫剧管线(可灵 / 即梦 / Vidu / Veo / Runway / LTX / 本地开源,或 LibTV / OiiOii / TapNow 这类成片 agent 平台)上手用。创作与出片仍在 skill / 外部工作台;工作台导出的 return-v1 包用 `gtrk ai-drama lay` 确定性回填独立 AI 轨。当用户想「上 AI 情景动画 / 做 AI 分镜 / AI 动画描述 / 把这段做成 AI 视频 / 给这段配 AI 情景动画 / 弗洛伊德这段怎么做视频」时使用本 skill。通用拆镜 craft(蒙太奇段分层 / 四段描述结构 / 中英双语 / 时长预算)由本 skill 自持;栏目视觉 DNA(Style Lock)从栏目配置 style.skills 里 produces==AI_DRAMA 的条目解析注入,绝不硬编某栏目风格。凡把口播派单里 AI_DRAMA 段落转成可出片的描述稿,优先用本 skill,别手搓。

7.7k tokens scripts zh
Gtrk Food Recap
by Gitruck

美食解说垂类图纸(解说链示例)——把探店/密着纪实长片(或自拍探店素材夹)与做饭流程记录,提炼看点后重述成一条中文美食解说成片工程。覆盖两形态:探店·密着(一家店/一顿饭/一个人的故事)与做饭流程解说(步骤线忠实重述)。菜谱知识类内容出局(那是配音链题材,走 gtrk-voiceover)。当用户想「做美食解说 / 探店视频解说 / 密着纪实翻讲 / 把这条做饭视频讲一遍 / 拉面店视频解说」时使用本图纸。链级条款(时序三档/引用段/出处对齐)以 gtrk-narration 为正本。

2.4k tokens zh
Gtrk Cover
by Gitruck

封面工作台——把一期视频的封面从「设计诊断 + 文生图 Prompt」到「H5 排字工作台一键导出多尺寸 PNG」两阶段做完。阶段一产诊断与三尺寸中英双版 Prompt(用户拿去 Midjourney / Flux / SDXL / Nano Banana / 即梦 / 可灵 / 通义万相 等任意文生图平台抽图),阶段二收齐背景图后产可拖拽排字的单文件 H5 工作台、浏览器里导出原分辨率 PNG。这是**纯创作 skill、无 gtrk 命令**(抽图在外部平台、排字在用户浏览器,同 /gtrk-ai-drama)。当用户想「做封面 / 出封面 / 封面 prompt / 封面排版 / 封面工作台 / cover」时使用本 skill。通用封面工艺(两阶段交付 / T1 字数布局路由 / Prompt 槽位 / 工作台引擎)由本 skill 自持;栏目封面审美(色板 / 字体 / 意象体系 / T1 样本)从栏目配置 style.skills 里 produces=="cover" 的条目解析注入,绝不硬编某栏目风格。凡给成片配封面,优先用本 skill,别手搓 prompt、别让用户裸排版。

3.2k tokens zh
Gtrk Long2short
by Gitruck

长剪短闭环——把一条长视频(课程/直播/播客/访谈毛片)通过同合云选段跳剪成多条高光短片工程,一次产出逐 clip 的 gtrk+剪映+PR 三方工程(可选智能分屏)。当用户想「长剪短 / 长视频切高光 / 直播切片 / 课程切短视频 / 播客出短片 / 两小时的课切成几条精华」时使用本 skill。凡涉及把长视频拆成多条可精修的短片工程,优先用本 skill 驱动 gtrk CLI 的 `long2short` 命令,不要手搓 ffmpeg、不要人工挑时间码。

3.1k tokens zh
Gtrk Live Slicing
by Gitruck

直播切片图纸——把一场数小时的直播回放,拆成一批可编辑的逐 clip 粗剪工程(gtrk+剪映+PR),交给用户自己挑、自己改、自己出片。含超长源片分段规约(服务端 2h 硬闸 + 60min 轮询墙钟)、选题清单确认停点、按画面形态配分屏、源片贴片如实告知。当用户想「直播切片 / 直播回放切高光 / 把这场直播剪成短视频 / 三小时直播怎么切 / 带货直播剪片段 / 讲座直播出短片 / 录播切片」时使用本 skill。凡涉及把直播回放拆成一批短片工程,优先用本图纸编排,别直接裸跑 long2short(超过 2 小时会被服务端直接拒)。

5.3k tokens zh
Gtrk Mg
by Gitruck

MG 动态图颗粒铺轨器——成片 SOP 第 ⑤ 步(**最后叠上**),在 B-roll 底轨三源(影视素材 / 本地素材 / AI 情景片段)全部铺齐、且 ④ 全局抽帧检查构图通过之后,把栏目的 html-particle 动态图颗粒(透明叠加 / 满屏底层)叠铺到已定稿的底轨之上。先由栏目 MG 生产 skill 按派单各槽位产颗粒,再驱动 `gtrk mg` 命令 lint + 铺进 `.gtrk` 工程的 beat_track。当用户想「铺 MG 颗粒 / 上动态图 / 铺动效 / 给这段配动画铺进工程 / 把 MG 派单铺轨 / 消费 dispatch.mg」时使用本 skill。凡涉及把 MG 动态图颗粒铺进已剪好的成片工程,优先用本 skill 驱动 gtrk CLI,别让用户自己去终端敲 `gtrk mg`。

14k tokens zh
Gtrk Matrix
by Gitruck

B-roll 检索铺轨编排手册——成片管线里第一个铺的车道(SOP ③ B-roll 底轨阶段)。消费拆分派单的 FILM_BROLL 队列(`dispatch.film_broll`),检索素材并在工程里平铺 N 条候选 B-roll 轨,供用户在 opencut 里用轨道小眼睛切换对比、挑选/调整;确认后交棒同阶段的 AI 情景片段腿(`/gtrk-ai-drama`,车道非空时)。命令族还含:**本地素材模式**(`matrix index` 建索引、`--local` 用自己的素材铺轨,**素材本体不上云**)、**按需理解**(`matrix describe`,为排序提供画质与看点信号)、**源时间窗检索**(`--source-window`,电影解说式逐段推进)、**通用三态素材检索**(`matrix material`:clip/image/audio 整条素材下载向,BGM/配图主场)、**plan 可编辑通路**(agent 改 plan 后 `matrix lay` 消费)。当用户想「铺 B-roll / 检索素材 / 找空镜 / 给空镜配画面 / 填 B-roll 候选 / 单独搜个词补个空槽 / 找 BGM / 搜配乐 / 搜张配图 / 用我自己的素材铺 B-roll / 给素材文件夹建索引 / 理解一下这些素材能不能用 / 把这堆素材剪成片 / 影视解说配画面」时使用本 skill。凡涉及把素材检索并铺进工程,优先用本 skill 驱动 gtrk CLI 的 `matrix` 命令族,别让用户自己去终端敲、也别手搓检索。

18k tokens zh
Gtrk Oralcut
by Gitruck

智能口播剪辑闭环——开剪前主动检查同目录独立口播音频,有则先对轨换声,再通过同合云剪掉废话/重复/长停顿,产出客户端(gtrk)+剪映+PR 三方工程。当用户想「剪口播/剪个视频/智能剪辑/去掉废话和停顿/把这条口播剪一版/出剪映草稿/生成剪辑工程/oral cut」时使用本 skill。凡涉及把口播毛片做成可二次精修的剪辑工程,优先用本 skill 驱动 gtrk CLI,不要手搓 ffmpeg。

4.9k tokens zh
Gtrk Narration
by Gitruck

通用解说视频创作图纸(解说链正本)——把「一个很长的东西」(旅行素材夹/影视长片/游戏实况/探店与做饭流程记录)提炼梗概、高光与看点,用精简叙述重讲成一条解说成片工程。判据=素材语料自带叙事时序且成片尊重它;声音基座无关(TTS 配音 / 用户自己念的口播毛片皆可,稿可 AI 代写可自备)。当用户想「做影视解说 / 游戏解说 / 探店解说 / 做饭流程解说 / 把这段素材讲成一条片 / 解说视频 / 把这部片子讲一遍」时使用本图纸。旅拍解说(gtrk-travel-recap)是本链先行垂类实例。凡解说类成片,优先用本图纸编排,别直接裸跑零件。

6.9k tokens zh
Gtrk Music Visualizer
by Gitruck

音乐可视化成片——把一首歌通过同合云出成一条频谱可视化视频(可选背景/封面 + 模板/配色样式)。当用户想「把这首歌做成可视化视频 / 音乐可视化 / 频谱视频 / 歌曲配画面出片 / music visualizer / 给歌配个动态封面视频」时使用本 skill。凡涉及把一段音频做成频谱可视化成片,优先用本 skill 驱动 gtrk CLI 的 `music-visualizer` 命令,不要手搓 ffmpeg。

1.5k tokens zh
Gtrk Ref Reading
by Gitruck

参考片精读——把一条别人的视频读成「可复刻」的整片模型与时码细节:逐字时码转写 + 带台词标签的抽帧网格 + 关键动效逐帧看 + 按「系统」记录每个设计事件对着哪句话,落成 ANALYSIS.md / TIMELINE.md / PROGRESS.md,读完可交给 /gtrk-style-maker 产栏目 skill,或直接指导把这套包装迁移到用户自己的片子上。驱动 `gtrk transcript --json`(取字级时码)与纯本地零计费的 `gtrk media` 看片命令族。当用户想「拆解这条视频 / 精读参考片 / 学一下这个博主的包装 / 分析这条视频的风格 / 这条视频的动画是怎么做的 / 照着这条参考片的风格做 / 把 XX 的风格蒸馏出来 / 参考片逆向」时使用本 skill。凡要从一条参考视频里学包装与节奏,先用本 skill 读透再动手,别看几张截图就开始仿。

3.6k tokens zh
Gtrk Style Maker
by Gitruck

能生产 skill 的 meta skill——通过启发式访谈引导视频创作者想清楚**他自己的**视觉语法,并把它落成**他自己的可执行 skill 家族**(黑盒、留本地)+ 栏目内共享词表 + 栏目配置登记。框架对审美零预设:不假设你有叙事结构、有容器概念、视觉分 MG/AI——你的维度(k)和取值(v)全部由你自己定义,本 skill 只提供「怎么想清楚」的工艺。当用户想「建一个新栏目/新频道的风格体系」「把我的审美沉淀成 skill」「做一个我自己的视觉 skill」「定义栏目风格规范/视觉语法」「栏目配置文档的 style 怎么填」「给我的栏目造一套创作工具」时使用本 skill。产出的 skill 通过通用 skills 适配器安装到当前或多个 Agent,登记进 gtrk 栏目配置(style.skills)。

3.3k tokens zh
Gtrk Talking Head
by Gitruck

通用口播视频创作图纸(口播链)——把一条或一组真人出镜的口播毛片,做成完整包装的可出片工程:外录音轨对齐换轨 → 多段拼接 → oralcut 粗剪 → 拆分派单 → B-roll/MG 字卡 → BGM → 字幕 → 客户端出片。当用户想「口播视频从毛片做到成片 / 口播完整包装 / 我拍了几段口播帮我做成一期 / 领夹麦音轨怎么合进去 / 口播加 B-roll 加字卡 / talking head 成片」时使用本 skill。凡涉及真人口播的完整成片编排,优先用本图纸驱动全链,别只跑一句 oralcut 了事、也别手搓 ffmpeg。

4.3k tokens zh
Gtrk Splitter
by Gitruck

视觉拆分派单器——把一条已剪好的口播工程(gtrk + transcript)拆成 beat 级「视觉拆分稿」,为每个连续文稿段指定唯一主层(A_ROLL 真人出镜 / MG 动态图 / AI_DRAMA 再现 / FILM_BROLL 影视素材)+ 辅助层,产机器可消费的派单清单(MG 槽位 / B-roll 检索队列 / AI 动画队列)驱动后续铺轨。当用户想「文稿视觉化拆分 / 视觉拆分稿 / beat 时间线 / 派分镜 / A-roll B-roll 分配 / 哪些段做动态图·哪些做 AI 情景动画·哪些用影视素材 / 给这条口播派单」时使用本 skill。凡涉及把口播成片拆成分镜派工,优先用本 skill 驱动 gtrk CLI 的 `split` 命令,绝不手写时码、绝不抄原文定位。

7.1k tokens zh
Gtrk Transcript
by Gitruck

将用户本地视频或音频(自备配音)转成一个飞书妙记式 Markdown 文字稿,并由 Agent 基于完整转写生成总结;音频配 --json 时另产句级时码 transcript.json(音频驱动工程兜底路输入)。用户说“视频转文字稿 / 视频转文字 / 提取视频文稿 / 把本地视频整理成文字稿或妙记 / 给我自己录的配音出时码稿”时使用。只处理本地文件,不接受 URL、平台链接或远端下载。

1.8k tokens zh
Gtrk Vlog Docu
by Gitruck

Vlog 纪实类一站式成片图纸——纪实 Vlog 是散文的载体:用有限的素材,把一个故事、或一个几个思想,串联起来(形散神不散)。表达形态上是「现场同期声 × 后期旁白」双声道交替,区别于纯口播链(单一对镜说)与纯配音链(全片旁白配 B-roll):同期声片段是内容本体不是配画,旁白只做定调层。当用户想「做一条纪实 vlog / 我拍了一堆素材想做成有深度的 vlog / 探店探险探厂记录 / 日常生活流做成片 / 旅途记录 / 野采野钓记录 / 城市漫游 citywalk / 圣地巡礼 / 取景地探访 / 挑战企划记录」时使用本 skill。凡涉及把带同期声的现场素材做成纪实成片,优先用本 skill 编排 gtrk CLI 全链,不要手搓单命令、别让用户逐步驱动。

8.3k tokens zh
Gtrk Tools
by Gitruck

gtrk 单点工具与媒体转换能力的调用向导,覆盖 `gtrk tool` 家族。当用户想“让图动起来 / 抠像去背景 / 图片或视频去黑边 / 图片或视频比例转换 / 视频防抖 / 添加蒸汽波滤镜 / 清理有权处理的图片或视频中的水印、字幕或叠加元素 / 视频超分变清晰 / 视频补帧变流畅 / 分离人声和伴奏 / 提取伴奏 / 按说话人分轨 / 音频变调变速 / 钢琴扒谱转 MIDI 或修复钢琴录音 / 给音频或视频降噪 / 去掉音频静音停顿 / 长图转方图 / 图片变 LivePhoto 微动 / 视频机械分镜或智能分镜切镜头 / 提取运镜高光片段 / 给视频加字幕或翻译字幕或烧录字幕 / 把视频翻译配音成另一种语言或译制 / 一键剪 MAD / 用某个 gtrk 单点工具 / gtrk 工具族里有什么”时使用本 Skill。凡涉及这些能力,优先驱动 CLI,不让用户自己去终端敲,也不手搓 ffmpeg。

9.6k tokens zh
Gtrk Voiceover
by Gitruck

通用配音视频创作图纸(配音链快速成片预设)——写好的稿子(或让 AI 代写)→ 配音 → 自动配画面 → 字卡/BGM/字幕 → 客户端可出片工程。适用素材为叙事服务、无自带时序的题材:知识科普、情感电台、观点输出、带货文案、盘点合集。当用户想「拿这篇稿子出条片 / 帮我写稿配音做成视频 / 科普视频 / 情感文案视频 / 把文字变成视频」时使用本图纸。素材自带时间线的解说类(旅拍/影视/游戏/探店)走 gtrk-narration,别用本图纸硬套。

2.2k tokens zh
Gtrk Travel Recap
by Gitruck

旅拍解说一站式成片图纸——输入一个旅拍素材文件夹(长视频或素材集),AI 理解素材、写三段式解说稿、一次确认后全自动跑完配音/建工程/拆分/B-roll/字卡/BGM/字幕,直达客户端可出片工程。当用户想「旅拍解说 / 旅行视频解说 / 把这条旅行长视频做成解说视频 / 旅拍素材做成片 / 长视频二创解说 / vlog 解说化 / travel recap」时使用本 skill。凡涉及把旅拍素材(自拍或网络长视频)做成中文解说成片,优先用本 skill 编排 gtrk CLI 全链,不要手搓单命令、别让用户逐步驱动。

11k tokens zh
Accessorysetupkit
by dpearson2699

Discover and configure Bluetooth and Wi-Fi accessories using AccessorySetupKit. Use when presenting a privacy-preserving accessory picker, defining discovery descriptors for BLE or Wi-Fi devices, handling accessory session events, migrating from CoreBluetooth permission-based scanning, or setting up accessories without requiring broad Bluetooth permissions.

3.5k tokens

Claude Skills — questions

Answers built from the skills we actually parsed.

What is a Claude Skill?
A folder with a SKILL.md file: instructions that teach an agent to do one thing well, optionally with scripts and reference files alongside. The format is open and called Agent Skills — Claude Code, Codex and other agents read the same files. It is not a program you run; it is knowledge the agent loads when the task calls for it.
How is a skill different from an MCP server?
A server gives the agent new abilities — it connects to something and exposes tools. A skill gives the agent knowledge: how to use what it already has. They combine, and often literally: 13 478 of the skills here declare which MCP servers they need to work.
Why are there fewer skills here than in other catalogues?
Because we deduplicate by content. Of 86 961 files found on GitHub, 68 678 are unique — the rest is the same skill copied into someone else's repository, word for word. Catalogues that count files rather than skills show every copy as a separate entry.
What does the token count mean?
A skill is loaded in stages. Its name and description stay in the model's context all the time; the body of SKILL.md is read when the skill is used; other files in the folder, such as references, scripts and fonts, are opened only when the skill asks for them. The number we show is SKILL.md, the part every use pays for. A big folder is not a big cost: one official skill ships 83 font files, and none of them ever enters the context.
How do I install a skill?
Copy the skill folder into ~/.claude/skills for personal use, or into .claude/skills inside a project. The agent picks it up by the name in the SKILL.md header — which is worth checking: 8 467 skills here share a name with another skill, and two of them cannot sit side by side.