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.
47,307 skills from 2,272 repositories · updated 9 hours ago
10,066 skills
- Slashbooksagentskillexchange/skillsReplace QuickBooks with an AI agent you control: import bank and credit card activity, categorize and reconcile transactions, close the month, and export the files your accountant needs.OtherNo warnings★ 58↓ —
- Ralph Loopagentskillexchange/skillsRun a ticket queue across agent sessions with a self-verifying loop: one item per iteration, verification gates outside the worker, commits only from a verified tree on a topic branch.OtherNo warnings★ 58↓ —
- eu-ai-act-high-risk-classifier-oliver-schmidt-prietzlawve-ai/awesome-legal-skillsDepth assessment of whether an AI system is high-risk under Art. 6 of the EU AI Act, grounded in the Commission's draft Art. 6(5) classification guidelines (general principles + Annex I + Annex III). Covers the Annex I product-safety route, all eight Annex III areas with worked examples, the Art. 6(3) exception and its profiling re-exception, and the Art. 25 quasi-provider trap. Outputs a structured decision block, a practitioner memo, and a JSON interchange artefact.OtherNo warnings★ 905↓ —
- mock-enginealibaba/skill-upAn evaluation and evolution tool for Agent Skills.OtherNo warnings★ 1.4K↓ —
- skill-authoringuvwt/agentdock创建、设计、修改、重构和验证 AgentDock Skill 时使用；负责 Agent Skills 兼容的 SKILL.md、可移植核心、引用、辅助脚本、测试、安全边界和本地真实验证。OtherNo warnings★ 1.3K↓ —
- history-healthOutlineDriven/outline-driven-developmentUse when a user asks to audit what recall fed agents. Not for source or remote-system changes.OtherNo warnings★ 55↓ —
- giurisprudenza-lampolawve-ai/awesome-legal-skillsRicerca giurisprudenziale rapida per ottenere un quadro di orientamento iniziale su un tema giuridico. Si attiva con "giurisprudenza-lampo [tema]", "cerca giurisprudenza su [tema]", "sentenze su [tema]", "pronunce su [tema]", "orientamento giurisprudenziale su [tema]", o qualsiasi variante che chieda di cercare sentenze, pronunce, massime, orientamenti giurisprudenziali o precedenti su un argomento giuridico. Usa questa skill anche quando l'utente dice "cosa dice la giurisprudenza su", "ci sono sentenze su", "precedenti su", "come si è espressa la Cassazione/il TAR/il Consiglio di Stato su".OtherNo warnings★ 905↓ —
- Book to Mentoragentskillexchange/skillsConvert a book or long document into a reusable AI mentor with source-grounded lessons, guided practice, citations, and persistent learning records.OtherNo warnings★ 58↓ —
- rustc-basicsOutlineDriven/outline-driven-developmentUse when selecting RUSTFLAGS, configuring Cargo profiles, tuning release builds, reading assembly or MIR output, understanding monomorphization, or diagnosing compilation errors.OtherNo warnings★ 55↓ —
- Strix•开放重定向asdfgh1445/ctf-super-hubStrix 开放重定向测试手册,覆盖钓鱼跳转、OAuth 令牌窃取与白名单绕过;触发名:strix-open-redirectOtherNo warnings★ 839↓ —
- Ticket TriageWYRE-AI/msp-claude-pluginsVendor-agnostic PSA ticket triage: priority determination, categorization, routing, and initial response practices applicable to Autotask, ConnectWise, HaloPSA, and other platforms.OtherNo warnings★ 51↓ —
- mcp-vodsaahl/skills用于追剧/追番的技能，为AI提供搜索影视播放地址的能力，并支持在小米电视上直接播放。当用户想搜索影视、动漫、短剧、综艺等节目信息或更新进度时使用此技能。OtherNo warnings★ 164↓ —
- dpgen-simplifyjinzhezenggroup/computational-chemistry-agent-skillsPrepare, explain, validate, and run DP-GEN simplify workflows for reducing repeated or redundant DeepMD datasets. Use when the user wants to generate or modify `param.json` and `machine.json`, run `dpgen simplify param.json machine.json`, organize repeated simplify experiments, or inspect simplify outputs.OtherNo warnings★ 155↓ —
- 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↓ —
- Slack MCP Serveragentskillexchange/skillsAgent access to Slack conversations and workspace workflows.OtherNo warnings★ 58↓ —
- roblox-data-storesEL4CTEO/roblox-skillsPersistence and cross-server data - DataStoreService (UpdateAsync, retries, budgets, limits), session locking with ProfileStore, schema migrations, OrderedDataStore leaderboards, MemoryStore queues/sorted maps/hash maps, MessagingService, right-to-be-forgotten. Use when saving player data, building leaderboards or queues, or fixing data loss, duplication, or throttling.OtherNo warnings★ 52↓ —
- Check Agent Counterparty Trust with Agent Guildagentskillexchange/skillsChecks evidence about an unfamiliar autonomous agent before delegation, verifies signed Agent Guild passports, and returns a bounded hire, caution, or avoid recommendation without taking consequential action.OtherNo warnings★ 58↓ —
- ConnectWise Manage ProjectsWYRE-AI/msp-claude-pluginsConnectWise PSA project management: project lifecycle and status/type values, phases, templates, resource/team allocation, budgeting, billing methods, and project tickets.OtherNo warnings★ 51↓ —
- Strix•信息泄露测试asdfgh1445/ctf-super-hubStrix 信息泄露测试手册,覆盖报错、调试端点、元数据泄露与源码暴露;触发名:strix-information-disclosureOtherNo warnings★ 839↓ —
- huntress-incidentsWYRE-AI/msp-claude-pluginsHuntress incidents and the remediation lifecycle: querying incidents by organization and status, SOC-recommended remediation details, individual and bulk approve/reject, remediation execution status, and the ordering constraint that incidents resolve only after all remediations are processed.OtherNo warnings★ 51↓ —
- dsh-marker-readeralibaba/skill-upDecode the marker fixture used by the DeepSeek Harness integration example.OtherNo warnings★ 1.4K↓ —
- mcd-saver-strategisthujjun/mcd-saver-strategist麦门省钱精算师。当用户在麦当劳（麦当劳/麦麦/金拱门/麦乐送/到店取餐）点餐，并希望"最省钱、凑满减、把优惠券用到极致、性价比最高、在预算内吃好"时使用；也适用于"帮我看看这些券怎么用最划算""我想吃XX但想便宜点""30块钱能吃到什么"这类诉求。它通过麦当劳 MCP 拉取门店菜单、可用优惠券、营养信息，再用确定性组合优化算法穷举出总价最低的合规点餐方案，并在用户确认后下单。OtherNo warnings★ 1.5K↓ —
- gradle-wrapper-upgradegradle/gradle-skillsUse whenever the user asks to upgrade, bump, or update the Gradle wrapper, `gradlew`, `distributionUrl`, `gradle-wrapper.properties`, or the build's Gradle version, on Gradle 7.0–9.x — or asks *how* that is done. Covers both modes: 'upgrade the wrapper' means perform it; a 'how do I upgrade' question is answered with the copy-pasteable commands rather than by running them. Not for adding a wrapper to a build that has none.OtherNo warnings★ 53↓ —
- manage-safety-analysismatlab/simulink-agentic-toolkitCreate, populate, and manage safety analysis spreadsheets (FMEA, FHA, HARA, custom) and fault trees (FTA) in Safety Analysis Manager. Use when the user asks to perform FMEA, hazard analysis, fault tree analysis, safety analysis, or work with Safety Analysis Manager documents. Requires Simulink Fault Analyzer.OtherNo warnings★ 1.3K↓ —
fine-tuning-with-trl
KalarisLabs/research-agent-skills
OtherUpdated 3 days ago
From the author
Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need RLHF, align model with preferences, or train from human feedback. Works with HuggingFace Transformers.
Install
npx skills add KalarisLabs/research-agent-skills --skill fine-tuning-with-trl -a claude-code
Installs to .claude/skills/fine-tuning-with-trl/ · Each command fetches the skill straight from GitHub with the open skills CLI.
Download .zip
Byte-identical to the scanned commit, with the repository’s licence and a SOURCE.txt.
What it can touch
Instructions onlyUses the networkLicence: MIT
Safety checks on SKILL.md
No warnings.
Why 82
Security100
Provenance62
Maintenance100
Licence100
Adoption32
Files (5)
- SKILL.md
- references/dpo-variants.md
- references/online-rl.md
- references/reward-modeling.md
- references/sft-training.md