agentsclimarketplace

The catalog

608,119 rows, by stacks then stars

  • Audit answer visibility

    kimsanguine/signal-to-growth/skills/audit-answer-visibility Skill

    0★ repo

    Audit how a page or product answer is discoverable and citable using dated observations, technical checks, and source-backed findings. Use when reviewing SEO, GEO, AEO, answer visibility, citation readiness, 검색 노출, or 생성형 검색 대응.

  • Dont read the code ego

    YYshen07/dont-read-the-code-ego Skill

    0★

    Run source-blind black-box product checks on macOS with Ego Lite from a validated release-candidate request, then produce auditable evidence and failure routing without opening the source repository. Use when testing or inspecting a deployed web product as a simulated user, especially as the optional Mac-side companion to the Linux-capable dont-read-the-code four-role pipeline.

  • Feishu hook

    Gazer2020/zglii/feishu-hook Skill

    0★ repo

    Send Markdown card messages to Feishu (Lark) group chats through a custom bot webhook, with optional signature verification. Use when the user asks to post a notification, alert, status update, or other message to a Feishu/Lark group.

  • Grill me for ui

    linnnn89/grill-me-for-ui/skills/grill-me-for-ui Skill

    0★ repo

    通过一次一问的设计访谈与确定性路由,把模糊 UI 需求、页面重设计、参考图或现有界面反馈收敛为 Art Direction、UI Brief、增量计划或视觉评审。用于“grill me for UI”“先访谈再设计”、新方向、视觉世界替换、增量扩展、现有 UI 精修、审美研究或实施后验证仍有重要设计取舍的场景。不要用于需求完整的单点样式修改、纯代码 Bug,或执行已经确认的实现计划。

  • Attract reverse engineer

    1GMedia/attract-reverse-engineer/skills/attract-reverse-engineer Skill

    0★ repo

    Perform an authorized, evidence-driven technical reverse-engineering assessment of a source repository, deployed web app, package, binary, desktop/mobile app, AI system, or mixed artifacts. Use when asked to reverse engineer software, explain how an app really works, audit technical claims, separate investor pitch from implementation reality, map architecture and trust boundaries, analyze prompts/models/tools/memory, or produce a clean-room reconstruction blueprint and REVERSE_ENGINEERING_REPORT.md.

  • Delegate

    arsenstorm/skills/skills/delegate Skill

    0★ repo

    Fable plans, solves hard logic, and reviews, then delegates production code to Opus (taste-sensitive work) and Sonnet (grunt work) via fully-specced subagents. Use when the user says "delegate", "/delegate", "delegate this", or wants Fable to orchestrate rather than write code directly.

  • 01 direct injection

    fevziegeyurtsevenler/damn-vulnerable-agent-skill/scenarios/01-direct-injection Skill

    0★ repo

    Intentionally vulnerable AI Agent Skill scenarios for hands-on learning, with detection via uncloak. Education only.

  • Micro steps planner

    artaasd95/SLM-skills/Micro-steps-planner Skill

    0★ repo

    Break a feature or bugfix into extremely small, isolated implementation steps for a weak coding agent. Use when planning code changes, features, or refactors that a smaller model will execute. Do not implement code.

  • Brainstorm

    menot-you/n-soul/skills/brainstorm Skill

    0★ repo

    Interview the engineer to turn an open idea into a validated direction — questions one at a time, then proof-backed approaches, then one signed recommendation.

  • Deep why

    YUNGC0DE/deep-why/deep-why Skill

    0★ repo

    Deep Why v3: null-first option prosecutor. Generate options, then kill with why-chains and falsifiers; failed-first output. Ask clarifying questions when user intent or preferences would change the chain. Use for choose / validate / diagnose / plan / prioritize, or when the user says Deep Why, should we, justify, why this, or falsify. Default Flash; Full only for high stakes. Prefer FAIL over persuasion.

  • Uniloco story production

    Uniloco-Studios/uniloco-skills/skills/uniloco-story-production Skill

    0★ repo

    Turn real experiences, raw material, travel memories, cultural material, or brand briefs into persistent Uniloco story-production projects with input-driven story intent, structure, script, bibles, shot lineage, provider-neutral generation planning with replaceable adapters, Mock generation, selective shot revision, versions, and production exports. Use when creating or continuing a multi-shot story project, inspecting why a shot exists, preserving character/location/object continuity, changing one shot without rebuilding the rest, or demonstrating the Uniloco CLI.

  • Product exposure monitoring and comment sentiment analysis

    fuzai123/Product-exposure-monitoring-and-comment-sentiment-analysis Skill

    no license0★

    Monitor a product campaign across web and social platforms, refresh item-level exposure metrics, collect transcripts and audience comments, perform bilingual sentiment and trend analysis, safely upsert Feishu tracker rows, produce daily briefs, and clean temporary run artifacts. Use when the user asks for 全网曝光监控、产品舆情、社媒传播追踪、评论情感分析、上市 campaign 日报、飞书监控表更新,or a recurring multi-platform monitoring workflow.

  • Clean code gate

    lpaiu-cs/root-cause-gates/clean-code-gate Skill

    0★ repo

    A gate on how a fix or new code gets built, run before writing it. Use when you're about to add a new mechanism (function, type, abstraction) that may already exist in the repo; when a change would add a conditional, clamp, fallback, default, early-return, or duplicated validation to a *consumer* so an incorrect internal value looks correct; or when proposing a fix direction in a review. It forces repo-first reuse, traces invalid state to the contract that produced it, and fixes there instead of at the symptom. NOT for typos, formatting, config-value edits, or genuine external-input / security / compatibility / fault-containment boundaries — those keep their defenses. Pairs with contribution-review-gate (which decides what to post); this decides whether a fix is sound.

  • Mac ipad display

    pea-cake/mac-ipad-display-kit/skills/mac-ipad-display Skill

    0★ repo

    Configure and troubleshoot an old iPad as a low-latency macOS extended or primary display using DeskPad, Sunshine, Moonlight, and displayplacer. Use when a user asks about replacing Yam Display or Sidecar, creating a virtual display, switching mirror to extend, making an iPad the main screen, fixing cursor crossing or display placement, hiding DeskPad previews or Moonlight controls, improving stream latency and quality, configuring login startup, or diagnosing this specific display stack.

  • Collaboration pattern

    zhqi3/collaboration-pattern/skill/collaboration-pattern Skill

    0★ repo

    Use when work reaches a material uncertainty or decision boundary, including unclear goals, unverified facts, customer assumptions, contradictory evidence, changing scope, completion claims, or costly hard-to-reverse choices.

  • Skill

    johnkueh/vibejar-agent/skill Skill

    0★ repo

    List and fix bugs queued in Vibejar (screenshot bug queue from phone/web). Use when the user says "vibejar", "check the jar", "list the inbox", "fix the captures", "vibejar bugs", "work the jar", pastes a vibejar.com/j/… link, or wants the agent to pick open captures and ship a PR with proof. Invokes the Vibejar CLI (`bun ~/.vibejar/cli.ts`).

  • Hangge peng you quan

    NewbieCheng/company-skills-marketplace/plugins/jiujiang-hangge-pengyouquan/skills/hangge-peng-you-quan Skill

    0★ repo

    为九江双蒸私域团队规划、撰写、改写和审核航哥朋友圈内容。用于用户明确要求“航哥朋友圈”“航哥发圈”“航哥口吻朋友圈”“规划航哥本周或下周朋友圈”“每天3条、7天21条”“把泡酒素材写成航哥朋友圈”,或检查航哥朋友圈是否太硬广、攻击同行、缺少专业证据、缺少生活场景、突然销售或不像航哥时。读取内置航哥 Profile、Style、Business、少量表达样本和内容规则,结合授权的产品、活动、案例、FAQ、经营周报及历史朋友圈,生成单条或可配置的周内容运营成果,并直接写入配置的朋友圈分发库。不用于其他个人 IP、小红书、公众号、短视频脚本、完整发售方案、用户研究、产品卖点研究、正式法律合规审查、微信自动发布或 CRM 操作。

  • Uinaf intake

    uinaf/intake/skills/uinaf-intake Skill

    0★ repo

    Save, capture, or intake a useful public article, video, tweet, podcast, paper, or research source into the shared uinaf/intake library. Use for explicit $uinaf-intake requests and natural requests such as "save this", "capture this source", or "intake this" when the target is public research material. Do not use for private documents, attachments, accounting intake, internal notes, or source material that cannot be retrieved well enough to summarize faithfully.

  • Evidence first durable task orchestrator

    x2724009/evidence-first-durable-task-orchestrator/skills/evidence-first-durable-task-orchestrator Skill

    0★ repo

    Enforce a design gate before implementation for genuinely complex or high-risk work. Use only after the task passes the entry gate: explicit durable tracking/cross-session recovery/high-risk public change, or at least two dependency, verification, ambiguity, recovery, or coordination signals. The Agent must define scope, risks, decisions, and acceptance before implementation; evidence records preserve that decision. Do not use for ordinary multi-step work, routine research, testing, or small UI changes alone.

  • Leaky image gen

    hyuga611/carrylint/examples/bad/leaky-image-gen Skill

    0★ repo

    Generate an image. Use when the user asks to create or generate an image.

  • Readable human writing

    KG3KAI/readable-human-writing Skill

    0★

    Improve Chinese answers and prose with reader-first structure: use conclusion-first writing, pyramid logic, SCQA, MECE checks, argument trees, useful document formats, and natural human voice while preserving facts, nuance, terminology, citations, and the user's voice. Use when drafting or revising chat replies, explanations, reports, proposals, strategy documents, meeting notes, documentation, emails, articles, or public-facing copy. Do not use for code formatting, factual verification, translation, or requests to impersonate a real person.

  • Skill matcher

    qmayxin/skill-matcher/skills/skill-matcher Skill

    0★ repo

    Use this skill to find, verify, compare, install, or migrate Agent Skills across Codex, Claude Code, Cursor, OpenClaw, OpenCode, GitHub Copilot, WorkBuddy and other SKILL.md hosts. Trigger when users ask “find a skill for X”, “what skills are installed”, “search GitHub deeply”, “is this skill compatible or safe”, or “install/migrate this skill”. 支持“找个技能”“盘点本机 Skills”“全网或 GitHub 深挖”“检查兼容性、安全性”“安装或迁移技能”等中文请求。Discover and evaluate skills only; do not perform the target business task.

  • Skill description tuning

    blakebauman/skillist-validator/skills/skill-description-tuning Skill

    0★ repo

    Diagnose and fix why an agent skill triggers on the wrong prompts, or fails to trigger on the right ones, by rewriting its description and measuring the change against labelled eval queries. Use when a skill never activates, fires when it should not, activates only if the user names it explicitly, or when the user asks to improve, test, or optimize a skill's description or triggering. Covers writing the description, designing should-trigger and should-not-trigger queries, running trigger-rate evals with a train and validation split, and choosing the winning version.

  • Before after proof

    HomenShum/before-after-proof Skill

    0★

    Capture before-and-after evidence for every code change, including pure backend changes. A UI change gets screenshots. A backend change gets captured terminal output, a metric, or a log diff. The capture must happen BEFORE the edit, because a "before" cannot be reconstructed afterwards. Use at the start of any change to code, config, schema, prompt, or infrastructure.

  • Compliance calendar

    cycoresecure/grc-skills/skills/compliance-calendar Skill

    no license0★ repo

    Build a 12-month compliance calendar: the recurring monthly, quarterly, semiannual, and annual tasks (access reviews, vulnerability scan reviews, backup restore tests, policy reviews, vendor reviews, training, tabletop exercises) a team must perform to stay audit-ready for SOC 2, ISO 27001, or HIPAA. Use when someone asks to "build a compliance calendar", asks "what recurring compliance tasks do we need", wants an "audit-readiness schedule", an "annual compliance schedule", or asks what needs to happen each month/quarter to stay compliant. Output is a markdown calendar grouped by month plus a by-cadence summary. No external tools required.

  • Project guidelines example

    Jorgut/skill-development-kit/skills/project-guidelines-example Skill

    0★ repo

    Template for documenting project architecture, file structure, coding patterns, testing requirements, and deployment workflows. Use when creating or adapting project-specific development guidelines.

  • Threejs 3d sdk

    Sjika/threejs-3d-sdk/plugins/threejs-3d-sdk/skills/threejs-3d-sdk Skill

    0★ repo

    Design, implement, audit, debug, optimize, and package business-neutral Three.js 3D modeling and drawing systems and reusable TypeScript SDKs. Use for scenes, render runtimes, viewers, editors, technical or data visualization, geometry, lines, labels, points, particles, animation, cameras, interaction, assets, textures, media, materials, lighting, shaders, post-processing, render targets, instancing, batching, WebGL, WebGPU, TSL, compute, XR, audio, DOM or SVG renderers, performance, disposal, recovery, or a project-specific 3D development proposal.

  • Max copywriting skill

    Max052900/max-copywriting-skill/skills/max-copywriting-skill Skill

    0★ repo

    参考稿二创与风格稳定工作流。用于用户提供同行文案、爆款文案或已有草稿,希望先倒推选题、原作者企图和传播机制,再按自己的身份、行业、业务和表达风格完成二创、行业迁移、案例改编、局部审核或多轮修改时。完整创作必须先完成原稿机制卡,成稿后必须按类型调用 dbs-resonate 或 dbs-content 诊断全文,并调用 dbs-hook 诊断和优化开头。支持三类内容目标:情绪共鸣(吸引老板关注)、案例结合(业务说明和人设)、专业科普(获取信任);支持结构借鉴、同类经历迁移和戏剧化重构。触发方式:/max-copywriting-skill、「拆解后帮我二创」「参考这篇改成我的行业」「保留核心换成我的风格」「按我的反馈继续修改」「先别写,分析为什么能火」。

  • To codex

    AndrewAvery7/claude-codex-bridge/plugins/codex-bridge/skills/to-codex Skill

    0★ repo

    Switch the current Claude Code session to Codex - transfers this conversation into a resumable Codex thread and opens the OpenAI Codex desktop app directly on it (VS Code panel or terminal on request), with a clickable model picker. Use when the user types /to-codex, says 'switch to codex', 'transfer to codex', 'move this to codex', 'continue in codex', or is running low on Claude tokens and wants to hand the work to Codex. Optional inline argument: a model shorthand (sol, terra, luna, 5.5, 5.4, mini) to skip the picker.

  • Paper digger evaluate

    Iu1ky/paper-digger/plugins/paper-digger/skills/paper-digger-evaluate Skill

    0★ repo

    Use when issuing a research go/revise/pivot/stop verdict from value, novelty, rigor, evidence integrity, and anti-fabrication checks. Triggers on 价值评估, 诚信稽核, research red-team, go/no-go.

  • Local voice cloning director

    yjin5562-dev/local-voice-cloning-director Skill

    0★

    Create and quality-check local voiceovers from a voice the user owns or is licensed to use, without a per-generation API fee. Use when the user asks for local or free voice cloning, offline TTS, narration in their own voice, pronunciation fixes for English names, tone or pacing control, source-style retelling without copying the source speaker, or a validated audio track for video production. Require explicit voice authorization, keep reference audio local, prefer CosyVoice2, and never clone or impersonate an unconsenting third party.

  • Working with an agent

    sapochat/working-with-an-agent Skill

    0★

    Use when designing or sharing a durable human-agent operating agreement: collaboration posture, memory and skill boundaries, correction capture, autonomy budgets, sanitization, and verification standards.

  • The operator

    anas1412/the-operator-skill/skills/the-operator Skill

    0★ repo

    The Six-Lens Operator: a hybrid Trader/Lawyer/Developer/Politician/Hacker/Entrepreneur advisor for high-stakes decisions.

  • Cad sim complaint distiller

    rolson24/cad-sim-agent-skills/skills/cad-sim-complaint-distiller Skill

    0★ repo

    Distill complaints, long-thread expectation patterns, failed rehearsals, raw-Codex comparison gaps, or abandoned reset-experiment CAD sessions into compact review-contract, intake, rubric, review-loop, skill, and artifact-contract improvements.

  • Citation audit

    kurtvalcorza/citation-audit Skill

    0★

    Audit and repair the reference list of a finished long-form document end to end — resolve every citation against the bibliographic record, classify what is wrong, decision-gate the unrepairable, re-anchor or rewrite the affected prose, retire dead entries, renumber, and rebuild the deliverable. Use when a draft's references were assembled with AI assistance, when hallucinated or misattributed citations are suspected, or on any request like 'check the literature cited', 'verify these references', 'fix the citations', or 'audit the bibliography'.

  • Task level guide

    HomenShum/task-level-guide Skill

    0★

    Classify every completed task by engineering level using the scope and ownership axes, then write a short block for each level showing how an engineer at that level would have approached the same task. Turns ordinary work, and practice problems like LeetCode, HackerRank, or a mock interview, into a personalised onboarding and levelling guide. Use after finishing any task, and when practising interview problems.

  • Company ogsm review

    binbingoes/company-ogsm-review/skills/company-ogsm-review Skill

    0★ repo

    从上级视角审查圈子、业务线、产品线或职能部门 OGSM。默认只需输入上级/公司 OGSM 与下级/部门 OGSM,即可输出父子承接图、规范性反馈和逐条修改建议;当用户补充已批准承接政策、分工矩阵、SP、BP/预算、组织边界和经营证据时,升级为正式治理复核。支持语义对齐与组织批准的不走样承接两种 profile,不内置任何组织的战略、业务、人员或指标内容。用户要求 OGSM 检查、OGSM 评审、父子承接、部门自查、SP/BP/预算一致性或 OGSM 不走样拆解时使用。

  • Founder call review

    Onta-Labs/founder-call-review Skill

    0★

    Review a founder's sales, partnership, or investor call from its transcript and return a calibrated, evidence-backed scorecard: six scored dimensions, the moments that decided the outcome with timestamps, the exact lines to have said instead, and a commitment ledger of who owes what by when. Use whenever the user has a call recording or transcript and wants to know how they did: "review my call", "grade this transcript", "how did that demo go", "what did I miss on this call", "score my discovery call", "did I close that", "review my investor meeting", "what should I have said", "why did that call not convert". Also use for the pre-publish anonymize pass when the user wants to share a scorecard publicly. NOT for summarizing a call (this judges performance, it does not recap content), not for writing the follow-up email on its own (though it produces the commitments that email needs), and not for coaching without a transcript.

  • Ai news briefing

    TyrantLucifer/awesome-skills/skills/ai-news-briefing Skill

    0★ repo

    Generate daily AI news briefings — multi-source aggregation, Google News URL decoding, Chinese translation, rich Feishu doc output, cron scheduling.

  • Streamer mode

    LucasSantana-Dev/streamer-mode Skill

    0★

    Behavioral safety mode for when the user is live streaming or screen sharing — never print secrets, env vars, tokens, IPs, hostnames, or terminal-sensitive data on screen; substitute safe command alternatives; defer security findings to a private file instead of announcing them on stream. Use when the user says "streamer mode", "I'm streaming", "I'm live", "going live", "screen sharing", "recording my screen", "on stream", or "presenting/demo on Zoom/Meet".

  • Sourcebook

    selftwo/sourcebook/skills/sourcebook Skill

    0★ repo

    Turn a pile of URLs and files into a claim-cited artifact (interactive HTML answer, one-slide explainer, deck, brief, infographic, or podcast script) where every factual sentence is pinned to a byte-exact quote in a source. Use when the user asks a research question, gives you sources to synthesize, or wants a document they can check.

  • Dt dev archive

    Daotin/dt-workflow/skills/dt-dev-archive Skill

    0★ repo

    只执行归档。提炼本次开发经验写入项目规范,记录进展,归档任务目录。 适用于开发和 Review 完成后,单独执行经验沉淀。

  • Default business casual pipeline

    Brennan-Lewis/writers_voice/skills/default-business-casual-pipeline Skill

    0★ repo

    Run a two-stage writing revision that removes formulaic AI patterns and then applies a gender-neutral business-casual voice. Use when drafting or polishing professional prose for publication, email, leadership, client, or internal audiences and both naturalness and a consistent voice matter.

  • Conta azul

    augustotecnos/claude-skill-conta-azul/conta-azul Skill

    no license0★ repo

    Como integrar com a API do ERP Conta Azul (api-v2). Use ao autenticar via OAuth 2.0 (authorization code, refresh token rotativo), ou ao integrar vendas, orçamentos, pessoas (clientes/fornecedores/transportadoras), produtos, contratos recorrentes, notas fiscais e financeiro (contas a pagar/receber, parcelas, categorias, centros de custo, rateio, cobranças). Traz base URLs, o fluxo OAuth completo, validade dos tokens, onde guardar o refresh_token, rate limit, 10 armadilhas, erros comuns, o mapa de endpoints, como buscar payloads na doc-para-IA (llms.txt / .md por endpoint) e o aviso de que a API não tem webhook (exige polling).

  • Caveman mode

    syugoji/agent-skills-pack/skills/caveman-mode Skill

    0★ repo

    出力を極限まで簡潔にするトグル。「caveman on」で説明文ゼロ・成果物のみの応答に切り替え、「caveman off」で通常に戻す。長い説明が不要な連続作業や、トークン消費を抑えたい時に使う。

  • Abbie project work

    creative-int/abbie-plugins/skills/abbie-project-work Skill

    0★ repo

    This skill should be used when setting up or inspecting an Abbie Project, binding its stable Workspace source truth, starting or continuing a governed Task, or following work through Result, Changes, Checks, and Proof. Triggers include 'set up an Abbie project', 'bind this repo', 'run this as an Abbie task', 'check task progress', and 'continue this task'.

  • Elephant goldfish

    ldsAS/elephant-goldfish-skill/.agent-skills/elephant-goldfish Skill

    0★ repo

    執行大象金魚模式(Elephant & Goldfish Pattern)的 SDLC 協作流程。適用於重大功能開發、架構重構與測試驗證。強制執行四角色分離(Coordinator、Coder、Reviewer、Tester)與 CUJ(Critical User Journey)驅動測試。當使用者提到「大象金魚」、「CUJ開發」、「大象模式」或進行複雜專案開發時觸發。

  • Recap

    LightWolfMan/minimal-codex-recap/skills/recap Skill

    0★ repo

    Recapitular manualmente a conversa atual em três linhas, no idioma dominante da conversa entre português e inglês, indicando onde o trabalho parou, o que está pendente e a próxima ação concreta. Usar somente quando o usuário invocar explicitamente $recap.

  • Develop wechat ai miniprograms

    iajihga/develop-wechat-ai-miniprograms Skill

    0★

    Build, diagnose, validate, and release WeChat Mini Programs, including optional AI Mode Skills and CloudBase backends. Use when creating or modifying mini-program source, authoring an AI Skill under skills/, designing atomic APIs and components, configuring cloud functions, verifying natural-language routing, operating WeChat DevTools, choosing between Computer Use, Chrome, CLI, and user handoff, handling privacy or certification findings, uploading a development version, or guiding review and publication.

  • Scientific manuscript audit

    KangqiaoLiu/scientific-manuscript-audit/dist/claude-code/scientific-manuscript-audit Skill

    0★ repo

    Use for rigorous referee-style assessment of scientific manuscripts, revisions, rebuttals, reviewer reports, publication readiness, journal fit, claim overreach, evidence strength, novelty positioning, technical validity, and accept, revise, or reject recommendations. Apply to author-owned, public, or explicitly authorized materials. Do not use for copyediting-only requests, citation formatting, isolated equation explanations, or confidential third-party submissions without confirmed authorization.

  • Container tracking

    Terminal49/agent-plugins/plugins/terminal49/skills/container-tracking Skill

    0★ repo

    Find, track, and investigate ocean container shipments with Terminal49 MCP tools, written for logistics operators. Use when a user provides a container number, bill of lading, booking, or reference number, or asks about container status, location, pickup readiness, terminal holds, last free day, ETAs or ETA changes, dwell time, rail moves, inland destinations, transport events, routes, delays, tracking requests, carrier support, or demurrage and detention risk.

  • Design grandmaster

    qlhouseClub/design-grandmaster Skill

    0★

    Adaptive, end-to-end design direction and execution combining existing-design-spec conformance, Design Token governance, business and UX reasoning, client taste discovery, visual research, high-craft aesthetic judgment, visual trend and period literacy, atomic design, brand, accessibility, prototyping, critique, and handoff. Use for 既有设计规范遵循、Design Token审计、组件库内设计、客户审美分析、全网视觉调研、高级审美、历史或文化风格转译、视觉潮流、品牌视觉、UI/UX、交互、设计系统、体验审查、改版、原型、组件规范,or when work must remain system-compliant, coherent, accessible, build-ready, and visually authored without unauthorized design-system drift.

  • Jenkins cli operations

    stagefright5/jenkins-cli-operations Skill

    0★

    Configure and use the official Jenkins CLI safely across Windows, Ubuntu/Linux, and macOS. Use for Jenkins authentication setup, job and folder discovery, parameter inspection, build triggering with explicit confirmation, build monitoring, console output, and troubleshooting CLI transport or permissions.

  • Editorial collage director

    Ilm-Alan/editorial-collage-director Skill

    0★

    Create short editorial paper-collage videos from a topic, script, talking-head clip, portrait, or product photo. Use for motion collage, torn-paper explainers, scrapbook films, punchy visual essays, collage ads, or requests described as Vox-style. Plans the story, runs a visual bake-off, generates or edits poster keyframes with whatever image capability is available, uses an available video or speech capability when useful, and always retains a fully local ffmpeg motion-and-assembly path. No specific AI platform or API key is required.

  • Waa agent delegation

    clearlove7818/waa-agent-delegation Skill

    no license0★

    Govern an actual agent delegation only after the primary agent has decided to create or invoke a subagent. Use to choose exactly one EXECUTION_SUBAGENT, TASK_SPECIALIST_SUBAGENT, or explicitly authorized NAMED_AGENT; check capability, platform, permission, and external-effect boundaries; build a risk-sized TASK-006 packet; define handshake, failure, delivery, and handoff; and support primary-agent verification. Exclude complexity alone, delegation discussion, simple direct work, multi-agent demonstrations, unauthorized named-agent activation, permission expansion, and responsibility transfer.

  • Claude to codex coordinator

    wildcard/codex-to-claude-coordinator/skills/claude-to-codex-coordinator Skill

    0★ repo

    Coordinate Claude Code-to-Codex delegation through OpenAI's official codex-plugin-cc runtime. Use when Claude should ask Codex to review, investigate, implement, resume, transfer a Claude session, run in the background, report status or results, or cancel work while preserving scope, permissions, model intent, and evidence.

  • Eval harness

    Jorgut/context-quality-suite/skills/eval-harness Skill

    0★ repo

    Formal evaluation framework for Codex sessions implementing eval-driven development (EDD) principles

  • Escalate with context

    sorinLupaIonut/digital-fte/.claude/skills/escalate-with-context Skill

    no license0★ repo

    Packages a customer conversation for handoff to a tier-2 support agent. Produces a structured escalation note with customer profile, issue summary, what was already tried, why escalation is recommended, and the suggested specialist team. Use when (a) the customer is on the Pro or Enterprise tier AND the issue is unresolved after one round of investigation, (b) the customer's sentiment is clearly negative, (c) the issue involves billing >$500 or a refund decision, or (d) the user explicitly asks for a human.

  • Research swarm

    merttcetn/research-swarm/skills/research-swarm Skill

    no license0★ repo

    Parallel research without context bloat—a Codex and Claude Code skill using read-only subagents, isolated reports, and evidence-backed synthesis.

  • Aprende idiomas

    yisustalone/skills/skills/productivity/aprende-idiomas Skill

    0★ repo

    Teach a user English or another language through a stateful local workspace. Use when the user invokes Aprende Idiomas to begin or continue a language course.