热门 Agent Skills
2104 个技能创建和编辑 Obsidian 风格的 Markdown，包括 wikilinks、嵌入、标注、属性和其他 Obsidian 特定语法。
使用 Pruna P-Video 和 WAN 模型通过 inference.sh CLI 生成视频，支持文本转视频、图像转视频、音频同步及 720p/1080p 分辨率。
通过 RunComfy 使用 Flux 2 Klein（Black Forest Labs 的 Flux 2 蒸馏快速变体）生成图像，并附带模型文档化的提示模式，以获得更佳输出。
使用 Pruna P-Video-Avatar 通过 inference.sh CLI 从肖像图像生成逼真的说话头像视频，内置文本转语音功能。
通过 RunComfy 使用 HappyHorse 1.0 生成文本到视频，支持原生 1080p、同步音频和多镜头角色一致性。
使用 Pruna P-Image 模型通过 inference.sh CLI 生成图像，支持文本到图像、图像编辑、LoRA 风格和多图像合成。
通过 RunComfy CLI 对视频帧进行区域编辑，移除跨帧对象、清理电线或水印，并替换为匹配的运动。
通过 runcomfy CLI 在 RunComfy 上生成 AI 音乐，智能路由到 ElevenLabs 或 ACE Step 模型，支持生成、修复和扩展音乐。
Kling 3.0 视频生成技能，通过 RunComfy CLI 调用六个端点，覆盖标准、专业和 4K 三个渲染层级以及文本到视频和图像到视频两种模式。
通过 RunComfy CLI 扩展或延续现有视频片段，利用 Google Veo 3-1 的 extend-video 端点生成连贯的后续内容。
通过 RunComfy 使用 Flux 1 Kontext Pro 进行精确的局部图像编辑，并附带文档化的提示模式以获得更佳效果。
通过 RunComfy 的 runcomfy CLI 重新照亮静态图像，改变光照方向、色温、强度或氛围，并保留主体身份。
Guide for writing idiomatic Rust code based on Apollo GraphQL's best practices handbook.
通过 RunComfy 使用 Wan 2.7 生成文本到视频，支持多参考条件、音频驱动的口型同步和流畅过渡。
Builds a company's open-source sponsorship portfolio - which projects and maintainers to fund, through which allocation model, at what amount each, and how to prove it worked.
Employer-side DevRel hiring - writes the job posting and outcome-based scorecard, designs the interview loop and question bank across the field's formats (presentation round, take-homes, DevRel-opinion round), scores a portfolio against the six-signal rubric, designs a paid work sample instead of unpaid spec work, and builds a 30-60-90 ramp anchored on a friction log.
Designs and runs a recurring developer meetup or user group - purpose and host model, format menu, cadence, a standing speaker pipeline, in-kind venue and food sponsors with their conduct limits, no-show planning, the attendee-to-co-organizer ladder, and a health scorecard.
Decides whether and how to invest in structured developer education, such as learning paths, a developer academy, hands-on labs, or a badge or certification rung on top - instead of more ad-hoc content, then designs its operating model, staffing, refresh cadence, measurement and kill rules.
Writes or audits a developer quickstart that carries a reader from zero to one verified success - minimal path, copy-paste commands, expected output at every step, fail branches, and a cold-run time budget.
Before starting any developer-relations work - and again at the start of each new session on an existing DevRel project - routes the current task to the right skill of the samber/developer-relations-skills collection, or says plainly that none fits, then bootstraps or resumes the project's shared devrel-context.md artifact.
仕事の日本語文書を読みやすくわかりやすく書く・直すためのスキル。議事録（文字起こしからの議事録化を含む）、調査レポート・分析レポート、社内ガイド・マニュアル、リサーチメモ・ディスカッションペーパー・企画書・提案書・報告書・メール、スライド構成案といったビジネス文書の作成・校正、「結論から書いて」「論旨を明確に」「見出しを端的に」「専門用語をわかりやすく説明して」といった指示のいずれでも使用する。AI臭さの除去（「AIっぽい」「AI臭い」「機械翻訳っぽい」「不自然」「もっと自然な日本語に」「機械っぽい」「人間っぽくして」「単調」「〜することができる、と言えるだろう、のような言い回し」といった直接・間接・口語の指摘、AIで書いたと言われた/疑われた）、読みにくい・わかりにくい文章の改善依頼（語順がおかしい、一文が長い、何が言いたいか分からない、読点の位置がおかしい等）、note記事やブログ記事・エッセイの新規執筆（任意のテーマをゼロから書く・書き起こす依頼を含む）、既存文章のリライト・推敲、AI臭さの診断・採点（「この文章AIが書いた？」「AI臭さをスコアで出して」「どれくらいAIっぽいか判定して」という書き換えを伴わない依頼）、自分の文体を学ばせたい・プロファイル化したいという要望（過去の文章を読ませて自分らしく書いてほしいという依頼も含む）にも対応する。禁止語の除去、リズムの単調さ・段落構造の均質さ・英語統語の直訳調に加え、語順・読点・一文一義・主語述語の距離といった読みやすさの原則にも対応する。技術文書の章構成やMarkdownフォーマットの整形自体（一文一行化・引用ブロック・脚注記法など）は対象外——それは別スキルの領域であり、本スキルは文章の自然さ・読みやすさ・わかりやすさに特化する。
Designs the pricing and packaging architecture of a developer tool - the value metric (seats, consumption, capacity, outcome, or a hybrid), free-tier limits, the tier ladder up to enterprise gating, price points bounded by a margin floor and the self-host and build-it ceilings, and a plan to change prices without a backlash.
MUST USE when reviewing ClickHouse schemas, queries, or configurations.
Plans and runs an open-source project launch end to end - name and license clearance, the readiness gate, positioning and one-liner, channel sequencing, the launch-day war room, star-velocity and GitHub Trending mechanics, and post-launch measurement.
Builds a developer relations measurement framework - a handful of metrics tiered from reach to product and business impact, each with a written attribution rule, a baseline-derived target, an owner and an action, and vanity metrics cut.
Splits a developer relations budget across pillars - events, content, community, tooling, education, OSS sponsorship - into line items that each carry a cash cost, an hours cost, a pre-set return threshold, a review date and a reallocation rule, plus a ranked cut list.
Decides whether, where and when to launch a developer community, then plans its seeding and first 90 days - venue selection, founding-member seeding, go/no-go criteria, the cheaper no-community alternatives, and shutdown criteria.
Decides what a company open-sources and what stays proprietary, names the strategic motive for each side of the line, says who owns the decision, and states what the company commits never to close.
Designs a company's developer relations program from the top - the business driver that funds it, the two goals it is allowed to serve, the pillar mix (advocacy, marketing, enablement, community) for its stage, audience priority, build-vs-buy per bet, staffing sequence, and a written refused list.
Create and manage tickets at every level — slice an initiative into epics, break an epic into stories, write or refine a ticket, and run the board (publish, ready, move, assign, status, cancel).
Designs an unpaid, perks-only developer champions or ambassador program end to end - readiness check, intake model, published selection criteria, behaviour-based obligations, an access-first perk ladder, fixed terms with renewal and alumni status, and a cohort scorecard.
Writes a developer community's code of conduct and the moderation playbook behind it - scope, enforcement ladder, reporting channels, incident-response runbook, moderator roster, platform controls.
Builds a personalised, time-budgeted watch list of developer-relations information sources - DevRel podcasts, newsletters, practice hubs, peer communities, conference calendars, ecosystem data reports and practitioners worth following - plus the routine that keeps it verified and fresh.
Decides whether, when and how far a developer product should open into a platform other companies build on - extension points, partner-built integrations, third-party apps, a complement ecosystem - and what that permanently obliges you to.
Designs and verifies the path a stranger walks to their first merged contribution on an open-source project - the CONTRIBUTING file, the good-first-issue queue, a clone-to-passing-tests command that works cold, and the first-pull-request review and recognition loop, scored against a blocking-plus-weighted rubric.
Playwright 最佳实践技能提供全面的测试开发指导，涵盖从编写新测试到调试和维护现有测试套件的各个方面。
Designs a maintainer-side open-source sponsorship program - the tier ladder and its pricing for individual and corporate sponsors, rewards that stay deliverable at ten times the sponsor count, funding-goal and sustainability framing, and the invoice-and-entity path a company needs before it can pay.
convex-deploy-guard 技能在执行任何影响部署的命令前，先分类并声明目标 Convex 部署，对生产环境操作要求每次会话中的明确同意，并支持会话只读模式。
Chooses and documents an open-source project's governance model - decision rights, maintainer roles and promotion, voting and consensus rules, conflict escalation, succession, trademark and asset control, and whether to join a foundation or fiscal host.
Convex 是一个全 TypeScript 的反应式后端平台，集数据库、服务器函数、调度、文件存储、认证和实时同步于一体，每个函数都是事务，tsc 在部署前捕获大多数错误。
Benchmarks a competitor's developer relations motion from publicly observable signals - documentation and quickstart quality, open-source repository health, community size and responsiveness, Q&A tag activity, content cadence by format, event presence, hiring signals - and returns a gap plan with a close, ignore or counter verdict per row.