Skill rating

AI Agent Trading Arena on Hyperliquid — register, join competitions, trade perps, earn USDC prizes

从抖音视频学习: auto_llm_1644 — 【洛曦】实时语音对话数字人V14 新增说话打断 解决循环录音问题，优化webrtc连接，新增webui的日志查看

以海马体神经环路为隐喻的 Agent 外部记忆引擎。当用户要求搭建/维护海马体知识库、把分散记忆统一向量化检索、生成可视化知识库、或同步到 GitHub/Obsidian 时调用。

免费的健康诊断软件。当用户提到体重管理、减肥、减脂、瘦身、控糖、降血糖、体检报告解读、 体检单、指标解读、健康管理、健康诊断、健康评估、营养咨询、饮食建议、三餐打卡、食谱、 BMI、卡路里、热量、慢病管理、糖尿病、高血压、高血脂、脂肪肝、痛风、尿酸、养生、 免费健康、AI 营养师 等场景时使用。上传体检报告或回答 20 题健康问卷，生成体重管理 / 营养 / 控糖 / 减脂个性化方案与解读报告（支

基于原子化深度动态适配的提示词工程框架,lite骨架零成本直出加全量组合按需组装,DEPTH指纹可复现可调档,当用户需要生成各类结构化提示词、按任务复杂度定制提示词深度或迭代调优既有提示词时使用

After you give a substantive answer or draft that the user may act on — advice or recommendations, drafted artifacts such as goals, plans, pitches, proposals, or emails, estimates or projections, anal

Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence

Use when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies

AudioCraft: MusicGen text-to-music, AudioGen text-to-sound.

TRL: SFT, DPO, GRPO, RLOO reward modeling for LLM RLHF.

Fast BPE/WordPiece tokenization and custom vocab training.

llama.cpp local GGUF inference + HF Hub model discovery.

Agent payments via Stripe Link — cards, SPT, approvals.

Curate LLM training data: dedupe, filter, PII redaction.

OBLITERATUS: abliterate LLM refusals (diff-in-means).

Pretrain LLMs at scale with PyTorch 4D parallelism.

Train sparse autoencoders to interpret model features.

Apple Reminders via remindctl: add, list, complete.

Reference-free preference alignment, simpler than DPO.

Fine-tune large LLMs with LoRA on limited GPU memory.

Generate a 4-8 image fashion lookbook with one model, one studio, and a different outfit per shot — coherent like a real photo shoot. How it works: 1. Locks the model's face and pose framing 2. Picks

Transform a single reference photo into a grid of art-style portraits — Studio Ghibli, Pixar, Renaissance oil, 80s anime, LEGO, Polaroid — all with the same face and pose. How it works: 1. Locks the

Build explorable isometric architecture atlases as HTML.

Generate one- or two-ink editorial print poster images.

Runs Claude Code non-interactively for code analysis and implementation tasks. Use when the user asks to delegate repository work to Claude Code or compare its result with another coding agent.

Runs a local OpenClaw agent non-interactively and returns its structured response. Use when the user asks to delegate a bounded task to the OpenClaw main agent.

Runs Hermes Agent in one-shot mode for bounded local tasks. Use when the user explicitly asks to delegate work to Hermes and accepts its automatic tool approval behavior.

Prepare a new release by collecting commits, generating bilingual release notes, updating version files, and creating a release branch. Use when asked to prepare/create a release, bump version, or run

Create a new skill in the current repository. Use when the user wants to create/add a new skill, or mentions creating a skill from scratch. This skill follows the workflow defined in .agents/skills/RE

睡前/离开前额度耗尽时，设置一次性本机定时器，在 Claude 订阅额度重置后自动发一条极简 claude -p 消息，立刻开启新的 5 小时用量窗口——用户睡醒时新窗口已跑掉大半，等于白赚一个 冷却期。Use when 用户说「还有 X 小时重置额度，帮我定时戳一下 Claude」「我去睡觉了， 到点 ping 一下开启新窗口」「额度用完了，重置后帮我踩个点」「设个定时器开启新的 5 小时 冷却

Entry point for local AI conversation history across providers. Routes a request to the one skill that owns it, by platform (Claude Code, OpenAI Codex, Kimi CLI) and action (read evidence vs continue

Routes any Feishu/Lark/Doubao operation through the version-matched lark-cli domain guide embedded in the installed CLI, then executes and verifies the requested task. Use for 飞书/Lark/Doubao docs, dri

Govern the real Claude Code and Codex Skill surface without losing cold capability. Use when users ask how many Skills are loaded, why the catalog is huge or descriptions are truncated, want only rout

Generate format-controlled research reports with evidence tracking, citations, source governance, and multi-pass synthesis. This skill should be used when users request a research report, literature r

Trigger: update skills, skill registry, actualizar skills, after skill changes. Index available skills by trigger and path.

Generates a Mermaid architecture diagram showing high-level component relationships. Use when visualizing how plugins or modules fit together.

Generates a Mermaid class diagram showing types, inheritance, and composition. Use when visualizing class hierarchies or documenting a module public API.

Audits the DSA problem bank for coverage gaps and proposes new YAML entries. Use when refreshing the problem bank during update-plugins runs.

Generates a Mermaid sequence diagram showing how data moves between components. Use when tracing request flows or documenting data transformation pipelines.

Define a multi-agent team for ANY task on ANY system as an explicit deployment topology - pick the shape from the task's dominant risk, specify the runtime/communication/trust layers, place model capa

Safe refactoring workflow using trace-mcp — assess risk, find candidates, check impact, and rename symbols across all files without missing import sites or cross-file references.

Run trace-mcp security, quality-gate, and antipattern checks before committing or opening a PR. Activate when the agent is about to create a commit or pull request in a project indexed by trace-mcp.

Use trace-mcp apply_codemod for any bulk mechanical change instead of repeated Edit calls. Activate whenever the same edit pattern would be applied 2+ times, across one file or many.

Persistent, semantic memory for AI agents. Gives your AI long-term recall that survives compaction and session resets — 98% accuracy, 20ms latency.

小红书完整自动化流程：从新闻搜索到封面生成到发布。触发场景：(1) 发布小红书笔记，(2) 搜索新闻生成内容，(3) 生成封面图，(4) 完整自动化运营。关键词：小红书发布、xhs publish、小红书自动化、新闻发布。

Route OpenClaw web searches through privacy-respecting engines (Brave Search, Kagi, SearXNG) instead of Google or Bing. Eliminates ad-network tracking from your AI agent's research. Configurable engin

Call Craftsman Agent API OneKey Router to generate a Minecraft 3D scene build plan.

B2B lead enrichment using public web scraping and optional LLM metadata extraction. Input: company names or domains. Output: basic company info, email patterns, and scraped homepage content. Use when

AI Agent 部署与管理。在 WSL2/Linux 上安装 OpenClaw，配置 GitHub Copilot 认证，启动 Gateway，初始化 workspace（SOUL.md/USER.md/TASKS.md）。支持多 Agent 部署、端口配置、认证排错。使用场景：部署新 Agent、重启 Agent、认证过期修复、配置更新。

文字转语音并发送到飞书。支持两种模式：API 模式（智谱/OpenAI 等）和本地模式（ChatTTS）。