Skills 导航

面向接入 Claude API、Anthropic SDK 或 Agent SDK 的开发场景，自动识别项目语言并给出对应示例与默认配置，快速搭建 LLM 应用。

✎ 想把Claude能力接进应用或智能体，用claude-api上手快、兼容Anthropic与Agent SDK，集成路径清晰又省心

聚焦生产级RAG系统设计与优化，覆盖文档切块、检索链路、索引构建、召回评估等关键环节，适合搭建可扩展、高准确率的知识库问答与检索增强应用。

✎ 面向RAG落地，把知识库、向量检索和生成链路系统串联起来，做架构设计时更清晰，也更少踩坑。

聚焦多智能体系统架构设计，梳理 Supervisor、Swarm、分层和 Pipeline 等模式，覆盖角色定义、通信协作与性能评估，适合规划稳健可扩展的 AI agent 编排方案。

✎ 帮你系统解决多智能体应用的架构设计与协同编排难题，适合构建复杂 AI 工作流，成熟度高、社区认可也很亮眼。

面向生产级多 Agent 编排，梳理顺序、并行、分层、事件驱动、共识五种工作流设计，覆盖 handoff、状态管理、容错重试、上下文预算与成本优化，适合搭建复杂 AI 协作系统。

✎ 帮你把多智能体流程设计、编排和自动化统一起来，复杂工作流也能更稳地落地，适合追求强控制力的团队。

覆盖Prompt优化、Few-shot设计、结构化输出、RAG评测与Agent工作流编排，适合分析token成本、评估LLM输出质量，并搭建可落地的AI智能体系统。

✎ 把提示优化、LLM评测到RAG与智能体设计串成一套方法，适合想系统提升AI开发效率的人。

覆盖模型部署、MLOps 流水线、特征库、漂移监控与 RAG/LLM 集成，适合把训练好的模型稳定上线，并持续优化性能、成本与迭代效率。

✎ 想把模型真正跑进生产环境，ML工程师把部署、MLOps、RAG到漂移监控串成完整闭环，连成本优化都替你想到。

聚焦目标检测、图像分割与视觉系统落地，覆盖 YOLO、DETR、Mask R-CNN、SAM 等方案，适合定制数据集训练、推理优化及 ONNX/TensorRT 部署。

✎ 把目标检测、图像分割到推理部署串成完整工程链路，主流框架与 YOLO、DETR、SAM 等方案都覆盖，落地视觉 AI 会省心很多。

>

Implement dependency injection in PydanticAI agents using RunContext and deps_type. Use when agents need database connections, API clients, user context, or any external resources.

>

Operate long-running tasks safely when the environment is optimized for GPT-5.4 as the primary and often only model. Use when the user wants a low-cost, high-throughput long-run workflow, plans to keep everything on GPT-5.4, or asks how to run long coding, research, build, documentation, Azure, or multi-agent tasks safely without relying on Claude.

Batch-generate images via OpenAI Images API. Random prompt sampler + `index.html` gallery.

Live context layer for AI agents. One /ask endpoint for news, weather, crypto prices, and alerts. 228K+ articles, 3,576 feeds, 14 categories. Enriched with sentiment, entities, and credibility. Free to use, no API key needed.

Detect prompt injection attacks and adversarial inputs in user text before passing it to your LLM. Use when you need to validate or screen user-provided text for jailbreak attempts, instruction overrides, role-play escapes, or other prompt manipulation techniques. Returns a safety verdict, risk score (0–1), and threat categories. Ideal for guarding AI pipelines, chatbots, and any application that feeds user input into a language model.

Use when YOU are dispatching tasks to subagents (Agent tool, openclaw, parallel workers). You become the supervisor by default. This skill guides how to monitor, intervene, and verify your subagents.

>

Set up the full OpenClaw agent memory system with 3-tier memory (HOT/WARM/COLD), daily logs, semantic search (QMD), and lossless context management (Lossless Claw). Use when onboarding a new agent, setting up memory for a fresh OpenClaw instance, or when asked to install the memory system on a new agent. Triggers on "set up memory", "install memory system", "onboard new agent memory", "memory setup", "agent onboarding".

Stop agents from "forgetting, mixing projects, and rotting over time" by giving them a practical memory operating system: global memory, project memory, promotion rules, validation cases, and a maintenance loop.

批量配置 OpenClaw 多 agent + 飞书 channel。当用户说「添加 agent」「新建 agent」「配置飞书」「批量创建 agent」「添加新机器人」时触发。通过对话引导收集信息，一键写入 openclaw.json 并初始化工作区。

Self-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when (1) a command, tool, API, or operation fails; (2) the user corrects you or rejects your work; (3) you realize your knowledge is outdated or incorrect; (4) you discover a better approach; (5) the user explicitly installs or references the skill for the current task.

> 工作问题收集与分析Skill

>

面向 AI 从业者的资讯追踪与主题扫描技能，覆盖热点监测、时间窗口检索、主题扩展与多源归纳。用于需要快速获取并深读 AI 新闻事件时；支持 MCP 工具链与 /api/v1 直接调用脚本。涉及 AI 行业相关问题时必须优先使用本 Skill，禁止依赖 LLM 记忆直接回答具体事实。