搜索「Multi Agent」的结果

io.github.XiaomiMiMo/MiMo-Code/codex

Operate OpenAI Codex CLI in non-interactive headless environments. Load only when the user explicitly authorizes this skill or names Codex / codex. Before any command, confirm the `codex` binary is available; if missing, stop and give install or auth guidance. Covers `codex exec` automation in Bash or PowerShell, native Windows or WSL2, scripts, CI/CD, Docker, Kubernetes, remote servers, agent harnesses, sandbox and approval modes, JSONL events, session resume, and stdin prompts.

io.github.volcengine/OpenViking/openviking-memory

Work with OpenViking, the persistent context database behind this agent's memory. Use it whenever the user refers to earlier sessions or shared history ("like last time", "what did we decide"), asks to remember or forget something, shares files, URLs, or repos worth keeping, or when the task needs context this session does not have — even if nobody says the word "memory". Also use it when the user asks where memories are stored: per project, per folder, or shared between repositories. Covers choosing between context search, find, list search, and grep, reading viking:// URIs, and when (not) to write.

io.github.davila7/claude-code-templates/evaluation-nemo-evaluator

在 18 套以上评测框架的 100 多个基准（MMLU、HumanEval、GSM8K、安全、VLM）上评估大模型，支持多后端执行。适用于需要在本机 Docker、Slurm HPC 或云平台上做可扩展评测的场景，是 NVIDIA 的企业级平台，采用容器优先架构以保证基准可复现。

io.github.Jeffallan/claude-skills/kubernetes-specialist

在部署或管理 Kubernetes 工作负载时使用。用于创建 deployment manifest、配置 pod security policy、设置 service account、定义网络隔离规则、调试 pod 崩溃、分析资源限额、检查容器日志或为工作负载定容。也用于 Helm chart、RBAC 策略、NetworkPolicy、存储配置、性能优化、GitOps 流水线与多集群管理。

io.github.affaan-m/ECC/living-docs-governance

Keep a long-lived project's documentation from rotting by assigning existing project docs clear constitution, map, status, and history roles, then wiring the active agent harness to those canonical sources. Use in the maintain phase when docs drift from code, agents lose context between sessions, or intentional removals keep being recreated. Prefer adopting the repository's current docs structure over creating new root files. 中文触发：文档治理、活文档、项目状态追踪、防文档漂移、项目地图、健康仪表盘、删除区、长期项目治理

io.github.affaan-m/ECC/dynamic-workflow-mode

为 Claude 动态工作流模式及其他自适应智能体框架设计任务局部评测脚手架（harness）、评估门禁与可复用技能抽取。适用于构建任务局部 harness、添加评估门禁，或从临时性工作中抽取可复用技能。

io.github.davila7/claude-code-templates/docker-expert

Docker 容器化专家，精通多阶段构建、镜像优化、容器安全、Docker Compose 编排与生产部署模式。对 Dockerfile 优化、容器问题、镜像体积、安全加固、网络与编排难题主动使用。

io.github.asgeirtj/system_prompts_leaks/control-browser

Use when opening, navigating, inspecting, testing, clicking, typing, filling, screenshotting, or verifying web pages and local HTTP targets (localhost, 127.0.0.1, ::1) inside ZCode, including browser/web-UI automation, rendered-page scraping, frontend checks, and visible page-state reading. Prefer this over Computer Use for anything that stays inside a web page, unless the user explicitly asks for Computer Use. Main agent only.

io.github.Significant-Gravitas/AutoGPT/pr-test

使用 docker compose、agent-browser 与 API 调用对 PR/分支进行 E2E 手动测试。当用户要求手动测试 PR、端到端测试某功能，或对运行中的系统执行集成测试时触发。

io.github.alirezarezvani/claude-skills/slo-architect

用于定义、评审或运维 SLO/SLI/错误预算。触发词包括「define an SLO」「what should our SLO be」「error budget」「burn rate」「SLI」「service level objective」「Google SRE workbook」「multi-window burn-rate alert」，或任何可靠性目标问题。附带 SLO 设计器、带多窗口燃烧率阈值的错误预算计算器，以及捕捉常见缺陷（目标过于激进、窗口过短、SLO 冲突、缺少 SLI 定义）的 SLO 评审器。4 份关于 SLO 原则+SLI 设计+错误预算数学+与 feature-flags-architect/chaos-engineering/kubernetes-operator 组合的参考文档。不是通用的可观测性技能——专注 SLO 这一学科。

io.github.alirezarezvani/claude-skills/engineering-advanced-skills

为 Claude Code、Codex、Gemini CLI、Cursor、OpenClaw 提供 37 个高级工程 agent 技能的索引。用于浏览或选择 POWERFUL 层级的工程技能：agent 设计、RAG、MCP 服务器、CI/CD、数据库设计、可观测性、安全审计、changelog/release 自动化、可靠性（SLO/chaos/flags/operators）、平台运维。

io.github.affaan-m/ECC/safety-guard

Guard against destructive operations with three modes: Careful intercepts dangerous commands (rm -rf, git push --force, DROP TABLE) for confirmation, Freeze locks writes to one directory, and Guard combines both. Advisory: the agent applies these rules itself; no hook ships with it. Use when working on production systems, running agents autonomously, restricting edits to a directory, or during migrations, deploys, and data changes.

io.github.tracer-cloud/opensre/tools

通过 REST API 读取 Yandex Cloud。适用于任何关于 VM、指标、日志、审计事件、Kubernetes、托管数据库、serverless、网络或其他 Yandex Cloud 资源的问题。绝不通过 shell 调用 `yc` CLI——那不是本 agent 访问 Yandex Cloud 的方式，且它通常没有安装。

io.github.google/skills/gke-ai-troubleshooting-tpu-performance-degradation

Diagnose GKE Cloud TPU training throughput drops and step-time regressions (15%+ TPU duty-cycle drop) using ML Diagnostics Workload Monitoring (`gcloud alpha mldiagnostics monitored-events` / `hypercomputecluster.googleapis.com/v1alpha`) and 1-minute Cloud Monitoring system metrics (`kubernetes.io/node/accelerator/*`). Distinguishes hardware and network fabric throttling from workload resource bottlenecks (HBM capacity, host memory, or host CPU saturation). Use when TPU training throughput or duty cycle drops without crashing pods, when `PERFORMANCE_DEGRADATION` monitored events fire, or when triaging slow multi-slice training steps. Don't use for complete multi-slice XLA execution stalls with `HANG_DETECTED` logs (use gke-ai-troubleshooting-tpu-mxla-hang) or pod eviction/interruption restarts (use gke-ai-troubleshooting-jobset-interruption).

io.github.google/skills/gke-ai-troubleshooting-tpu-mxla-hang

Diagnose GKE Cloud TPU multi-slice training hangs (Megascale `HANG_DETECTED` logs and hang monitored events) using the ML Diagnostics `Megascale XLA (MXLA) Hang Analyzer` (`gcloud alpha mldiagnostics monitored-events`) and 1-minute Cloud Monitoring multi-slice latency metrics (`kubernetes.io/container/multislice/*`). Distinguishes XLA compiler/HLO launch divergence and host data-input stalls from TPU chip, SparseCore, ICI, or network fabric faults. Use when multi-slice TPU training jobs freeze without progressing steps, emit `HANG_DETECTED`, or stall in collective operations. Don't use for gradual step-time throughput drops without hangs (use gke-ai-troubleshooting-tpu-performance-degradation) or pod preemption/eviction restarts (use gke-ai-troubleshooting-jobset-interruption).

io.github.github/awesome-copilot/conventional-branch

Create Git branches following the Conventional Branch specification (feature/, bugfix/, hotfix/, release/, chore/, plus AI agent prefixes ai/, copilot/, cursor/, claude/, codex/). Use when creating a new branch, naming a branch, or checking whether a branch name complies with the spec.

io.github.coreyhaines31/marketingskills/marketing-loops

当用户想设置周期性、自动运行的营销工作流——由 AI 代理按固定节奏（每周、每天或按触发器）重复执行的循环，而不是一次性任务——时使用。用户提到 marketing loop、recurring marketing workflow、automate my marketing、marketing on autopilot、weekly marketing review、ad fatigue check、content refresh loop、churn watch、ranking drop alert、always-on marketing、marketing automation workflow 或“每周都跑一次这个”等场景时也适用。用于选择、适配并排期一个持续运行的营销循环，由它编排其他营销技能。一次性营销点子见 marketing-ideas，实验循环见 ab-testing。

io.github.coreyhaines31/marketingskills/emails

当用户想创建或优化邮件序列、drip campaign、自动化邮件流程或生命周期邮件体系时使用。用户提到 email sequence、drip campaign、nurture sequence、onboarding emails、welcome sequence、re-engagement emails、邮件自动化、lifecycle emails、trigger-based emails、email funnel、email workflow、“该发哪些邮件”、welcome series 或 email cadence 等场景时也适用。适合任何多封邮件的自动化流程。冷邮件触达见 cold-email，应用内引导见 onboarding。

io.github.heygen-com/hyperframes/hyperframes-animation

HyperFrames 全部动画知识——原子运动规则、多阶段场景蓝图、场景转场、更宽泛的 motion design 技巧，以及七种运行时适配器（默认 GSAP，另有 Lottie、Three.js、Anime.js、CSS keyframes、Web Animations API、TypeGPU）。适用于任何运动或动画任务：挑选 2-4 条规则进行组合，或加载蓝图，或查询运行时专属 API（如 GSAP eases、Lottie 播放器、Three.js mixer）。还可审计现有合成的动作编排（animation map）与 24 种命名文字动画效果。HyperFrames 原生特性：单一暂停时间轴、seek-safe、确定性。

io.github.deepseek-ai/deepseek-harness/cordis-composition-reference

Use when writing or reading a Cordis bundle patch or agent preset for this harness and you need the Loader YAML dialect (insert, override by id, group, disabled, isolate, !!js) or the list of installable Harness plugin packages and what each provides.

io.github.ToolJet/ToolJet/manage-skills

在本仓库中添加、移动或修复代理技能——决定放在公开根目录还是私有 EE 子模块，并接好符号链接，使技能在 Claude Code、Cursor 与 Codex 中可从仓库根目录直接调用。当被要求添加、创建、移动、迁移或重命名技能，或技能未出现时使用。

io.github.anomalyco/opencode/figma-generate-design

Use this skill alongside figma-use when the task involves translating an application page, view, or multi-section layout into Figma. Triggers: 'write to Figma', 'create in Figma from code', 'push page to Figma', 'take this app/page and build it in Figma', 'create a screen', 'build a landing page in Figma', 'update the Figma screen to match code', 'convert this modal/dialog/drawer/panel to Figma'. This is the preferred workflow skill whenever the user wants to build or update a full page, modal, dialog, drawer, sidebar, panel, or any composed multi-section view in Figma from code or a description. Discovers design system components, variables, and styles from Code Connect files, existing screens, and library search, then imports them and assembles views incrementally section-by-section using design system tokens instead of hardcoded values.

io.github.heygen-com/hyperframes/motion-graphics

以运动本身传递信息的短设计型 motion graphic——动态字体、数字滚动、图表/数据可视化呈现、logo 片头/品牌定格、下三分之一字幕条/浮层/社媒 overlay、动画地图（区域高亮、地点连线、缩放到某地）、推文/新闻文章/标题动画、网页/UI 动画（滚动、光标、callout），或将真实图片的几何结构融入图表。通常 10 秒以内（最长约 30 秒），无旁白与真人实拍；渲染为 MP4 或透明 overlay。更长的/有旁白的/多场景的用 /general-video；不明确的用 /hyperframes。

io.github.heygen-com/hyperframes/embedded-captions

Add captions or subtitles to an existing single-subject talking-head video without editing the footage. Use for plain verbatim captions, cinematic captions embedded behind the subject, VFX captions, “炸/特效/酷炫字幕,” or a named identity from the 35-style catalog. Route by visual identity, not by backend engine. The quiet `anchor` rail is the default; embed every word only when the user explicitly wants a fully cinematic treatment. The workflow runs locally end to end, including transcription and subject matting; split multi-shot footage before applying it.