Agent skills you can check before you install
Public SKILL.md skills from GitHub repositories with 50+ stars and from the vendors themselves, scanned and scored. Install any of them with one command in Claude Code, Cursor, Codex, GitHub Copilot or Gemini CLI.
46,095 skills from 2,198 repositories · updated 56 seconds ago
9,543 skills
- dsh-plugin-creatorlovstudio/skillsCreate a @deepseek-ai/dsh-* plugin package end-to-end — choose the extension point or capability seam, scaffold the package, implement the tool/hook/service, and run the repo gates. 触发：新增插件 / 加工具 / 开发 capability。OtherNo warnings★ 73↓ —
- motherduck-load-datamotherduckdb/agent-skillsLoad files, object storage, dataframes, or external databases into MotherDuck with bulk ingestion. Use for any import of data into MotherDuck.OtherNo warnings★ 62↓ —
- praise-before-worklovstudio/skills落实用户“每次帮我肝活之前先夸夸我”的个人协作偏好。新任务开工前，先明确、具体地赞美用户，再开展工作。显式调用本 Skill 时，即使用户只说“你好”也要给出能明显感受到的夸奖；普通进度更新不重复。OtherNo warnings★ 73↓ —
- go-rust-reversecoco-research/cocoUse for reverse engineering stripped Go and Rust binaries including runtime recognition, pclntab/moduel data recovery, panic strings, and idiomatic decompilation recovery.OtherNo warnings★ 795↓ —
- Board HygieneWYRE-AI/msp-claude-pluginsBoard-wide ticket maintenance, distinct from working any single ticket: stale-ticket detection with per-status staleness thresholds, the Waiting-on-Client rot case and its three valid resolutions, duplicate/related-ticket detection signals and safe linking, and technician queue-balance assessment weighted by more than raw ticket count.OtherNo warnings★ 52↓ —
- Blumira MSPWYRE-AI/msp-claude-pluginsBlumira's MSP path group (`/msp/*`): managed-account enumeration, cross-account and per-account finding queries, per-account device, agent-key and user management, and how MSP paths differ from org paths.OtherNo warnings★ 52↓ —
- protocol-reversecoco-research/cocoUse for authorized reverse engineering of custom binary protocols, Protobuf/gRPC, WebSocket frames, and PCAP-driven protocol recovery.OtherNo warnings★ 795↓ —
- cachingnahid-sparktales/agent-dispatcherDecide what to cache, at which layer, for how long, and how it gets invalidated — before any cache is added. Use when someone proposes caching to make a read path faster, when stale or cross-user data is suspected, when adding a TTL or a Redis/CDN layer, or when reviewing a change that introduces one. Not for fixing a slow query or an N+1 (do that first — a cache over a bad query hides it), and not for HTTP header tuning with no server-side store.OtherNo warnings★ 59↓ —
- no-commentsdsebban/skillsSpawn Comment Sicko, fix accepted findings, and offer encodings for claimed constraints.OtherNo warnings★ 69↓ —
- blender-uv-texturingCheshireJCat/blenderUV unwrap, atlas-map, project textures, use alpha decals, bake maps/lightmaps, and prepare texture-driven Blender assets for glTF/GLB export. Use whenever the user provides texture packs, texture atlases, decals, UV layouts, lightmaps, wants a texture to fit a mesh 1:1, or reports stretched/off textures. Pairs with reference-to-3d for template-accurate reconstruction and blender-materials for PBR values.OtherNo warnings★ 63↓ —
- runZero AssetsWYRE-AI/msp-claude-pluginsThe runZero asset inventory: searching and browsing assets, asset attributes, OS fingerprinting, hardware details, and network interfaces.OtherNo warnings★ 52↓ —
- new-growing-products-in-categoryjoomcode/joompulse-skillsFinds new, already-selling listings in one category on JoomPulse — Mercado Livre (Brasil) or Shopee Brasil. Keeps listings that are recent (default: under about 30 days on air), well rated (above 3) and have real traction (about 30+ estimated monthly sales); all thresholds overridable. Returns a product table. Use when a seller wants fresh, promising entrants in a niche. Triggers: "new products in category", "what's launching and already selling", "recent best-sellers in my category", "new products on Shopee"; pt-BR "produtos novos na categoria", "novos anúncios que já vendem", "lançamentos em alta na minha categoria", "produtos novos na Shopee", "novidades na Shopee que já vendem". Ask which marketplace when unclear; never mix the two. Sales and revenue are JoomPulse estimates, not real transactions; price, rating and reviews are real. For well-selling but weak-rated products, use the high-demand low-quality skill; for one product over time, the change-monitor skill.OtherNo warnings★ 54↓ —
- blender-lightingCheshireJCat/blenderLight Blender scenes professionally — three-point setups, HDRI environments, studio/cinematic/dramatic configurations, light groups, color temperature, soft vs hard shadows. Use whenever the user asks to "light the scene", "set up lighting", "make it look cinematic / dramatic / studio / outdoor / sunset", "add a key light", "use HDRI", or any lighting-related request. Make sure to use this skill even if the user does not say "light" — also covers "make it look professional", "studio shot", "moody atmosphere", "golden hour", "rim light". Pairs with blender-materials (lighting reveals materials) and blender-cameras (lighting + composition together = shot).OtherNo warnings★ 63↓ —
- docusignAnil-matcha/awesome-muse-connectorsDocuSign eSignature: draft and send signature envelopes (demo env default), check envelope status, download signed documents. Trigger phrases: docusign, send for signature, signature envelope, envelope status, signed document.OtherNo warnings★ 1.4K↓ —
- deeplAnil-matcha/awesome-muse-connectorsTranslate text with DeepL: translate between 30+ languages, list languages, check usage. Trigger phrases: deepl, translate this.OtherNo warnings★ 1.4K↓ —
- Rootly ServicesWYRE-AI/msp-claude-pluginsThe Rootly service catalog: the real service record (ownership via owner teams/users, Slack channels, dependency modeling via linked service IDs, and third-party links to PagerDuty/Opsgenie/Cortex/Backstage/GitHub/GitLab), service CRUD, incident/uptime charts per service, and how services link back to incidents and alerts. Rootly has no built-in tier/criticality field -- that claim is removed, not renamed.OtherNo warnings★ 52↓ —
- xiaoyao-searchdtsola/xiaoyaosearch小遥搜索 MCP 工具 - 本地文件智能搜索（语义/全文/图像/语音/混合搜索）OtherNo warnings★ 1K↓ —
- verify-hypothesesTserenTserenov/FMT-exocortex-templateСверка журнала гипотез (hypotheses-log.md) по запросу вне ритма week-close. Фильтрует записи со статусом «на сверке» и наступившей датой, сверяет критерий фальсификации с доступной фактурой, предлагает вердикт, пилот подтверждает, вердикт пишется новой записью. Вызывать на явную фразу пилота, не автоматически.OtherNo warnings★ 63↓ —
- constitutional-aiKalarisLabs/research-agent-skillsAnthropic's method for training harmless AI through self-improvement. Two-phase approach - supervised learning with self-critique/revision, then RLAIF (RL from AI Feedback). Use for safety alignment, reducing harmful outputs without human labels. Powers Claude's safety system.OtherNo warnings★ 76↓ —
- minutes-tagsilverstein/minutesLightweight outcome tagging for meetings — won, lost, stalled, great, or noise. Use whenever the user says "tag this meeting", "mark that as a win", "that one was a loss", "tag yesterday's call as stalled", "mark this great", "that meeting was noise", "label that meeting", or any time they describe a meeting outcome in passing. Tagging takes 5 seconds and unlocks /minutes-mirror correlation analysis — the more meetings get tagged, the smarter mirror gets at telling the user what behavior patterns lead to wins. Surface this skill any time the user mentions a meeting result, win, loss, or wasted time.OtherNo warnings★ 1.6K↓ —Also in 1 other repo
- Self-Push: Lazy-Check Before Submittingagentskillexchange/skillsRun a structured lazy-check on any significant output before submitting. Forces root-cause analysis, alternative consideration, and outcome verification — especially critical in autonomous runs where no human reviewer will catch lazy shortcuts.OtherNo warnings★ 58↓ —
- deepline-enginegetaero-io/gtm-eng-skillsBuild, publish, and verify a Deepline Play as a durable state machine over Customer DB tables, including a small paid pilot. Invoke this skill explicitly when the user asks to build an engine.OtherNo warnings★ 72↓ —
- cipp-opsWYRE-AI/msp-claude-pluginsCIPP's own operational layer rather than the tenants it manages: GDAP role definitions and pending invites, the CIPP scheduler, and the ping/version/log endpoints used to diagnose why other CIPP tools fail silently.OtherNo warnings★ 52↓ —
- punct-polishtaxueseek/plainspoken-writing【标点打磨】AI写中文最常见的毛病就是标点：半角逗号、缺少中西空格、直引号、破折号滥用……这个技能一次性解决。覆盖中文散文、技术文档、中英混排、社交媒体四种场景，代码和URL完全不动，还具备英文标点校对能力。 触发：检查标点、修正标点、标点有问题、标点规范、半角改全角、中西空格、标点符号检查、punctuation、fix punctuation。 Not for：纯英文文本校对（→ 英文编辑工具）、纯代码文件、用户明确说"不要改标点"。OtherNo warnings★ 68↓ —
tribunal
arbiterForge/codeArbiter
OtherUpdated 4 hours ago
From the author
Run an opt-in deep codebase audit with persisted findings. Confirm cost before dispatch; filing and telemetry need separate approval.
Install
npx skills add arbiterForge/codeArbiter --skill tribunal -a claude-code
Installs to .claude/skills/tribunal/ · Each command fetches the skill straight from GitHub with the open skills CLI.
No download: this skill’s licence (AGPL-3.0) does not let us redistribute its files. The install commands fetch them from GitHub instead.
What it can touch
Instructions onlyNo network use seenLicence: AGPL-3.0
Safety checks on SKILL.md
No warnings.
Why 82
Security100
Provenance66
Maintenance100
Licence80
Adoption37
Files (24)
- SKILL.md
- references/ai-markers.md
- references/cost-and-models.md
- references/finding-record.md
- references/issue-filing.md
- references/lenses/appsec.md
- references/lenses/architecture.md
- references/lenses/change-closure.md
- references/lenses/coverage.md
- references/lenses/infra.md
- references/lenses/migration.md
- references/lenses/observability.md
- +12 more