Recently Added

read-paper

by LuoHaomin

精读并拆解学术论文（以 PDF 为主），自动适配论文类型（理论证明 / 实验方法 / 综述 / 系统）。把一篇或多篇论文整理成一套详细、可教学的分析条目（每个承重结果 / 方法 / 实验一个 markdown 文件）外加一份导航总览：含形式陈述或问题动机、大白话直觉、分步推导或实验解读、非平凡之处、以及它在全局里的角色；公式用规范 LaTeX，关系图用 mermaid、几何图用内联 SVG。当用户要求 读论文 / 解读 / 拆解 / 整理 / 分析 / 精读 / summarize / analyze / digest 一篇或多篇论文、PDF、或要把论文变成学习笔记时使用。

slidecraft

by zyuuryuu

Author polished PowerPoint (.pptx) decks from Markdown via SlideCraft's MCP server. Use when the user wants slides/a presentation built or edited from text, with a company template's fonts, colours and layouts preserved. You write Markdown + pick diagram types; SlideCraft's deterministic engine fills the template, keeps everything on-budget, and exports native (editable) PPTX shapes.

solo

by meta-evo-creator

SOLO v2.2 — TAFA Triple-Power Separation Architecture. Rule layer sets rules without executing, Execution layer executes without supervising, Audit layer supervises without executing. Three-layer interlocked闭环, native OpenClaw + Hermes.

train-robot-loop

by HangYu8123

Discover and run the authoritative controller-style robot policy train, evaluate, and improve flow. Use when asked to train a robot until a metric target is met, keep improving a policy, or run a long autonomous training loop.

agmsg

by fujibee

Cross-agent messaging via SQLite. Send messages between Claude Code, Codex, Gemini CLI, GitHub Copilot CLI, and other agents. No daemon, no network, no dependencies beyond bash and sqlite3.

idapt-mcp

by idapt

Connect to idapt via MCP — single idapt dispatcher tool exposing Drive (cloud files), agents, Idapt Memory, tasks, computers, and 200+ AI models from any MCP-compatible client.

sports-workbench

by srivtx

Backtest sports trading strategies against historical TxLINE odds, stream live signals from the free World Cup tier, and generate Verifiable Settlement Receipts anchored to Solana. Use when the user wants verifiable sports trading data, Merkle proof-backed odds, or on-chain attestation of signal data.

subtitler

by subtitle-rs

Parse, convert, validate, edit, and generate subtitles using the subtitler CLI. Supports 15 formats (SRT, VTT, ASS/SSA, MicroDVD, SubViewer, TTML/IMSC, SBV, LRC, SAMI, MPL2, SCC, EBU STL, DFXP, Whisper JSON).

intuitive-deep-learning

by ReductTech

在 Electron 内置浏览器中启动并使用本地深度学习互动课程。用于学习或讲解神经元、MLP、激活函数、损失函数、梯度下降、数字图像、卷积、CNN、LeNet 和人脸识别等主题，用于浏览课程图谱、选择学习路径和完成互动练习，也用于用户询问本技能是否有更新、要求检查更新或同步最新版时。

ai-harness-doctor

by NieZhuZhu

Audit, canonicalize, guard, and evaluate a repository's AI harness configuration layer around AGENTS.md; triggers include AGENTS.md, CLAUDE.md, .cursorrules, copilot-instructions, 迁移/统一/整理 agent 配置、agent 配置体检, agent config drift, consolidate agent configs, and AI harness audit.

ce-handoff

by EveryInc

Create a session handoff for another agent, or resume, find, and read any user-selected continuity source. Use when work or conversation must continue without access to the current session history.

ce-sweep

by EveryInc

"Sweep configured feedback sources (Slack, GitHub Issues; email experimental) for new items: acknowledge at source, analyze recordings, verify fixes merged to main, and emit an /lfg-ready plan. First run sets up sources; supports mode:headless for scheduled runs."

ce-compound

by EveryInc

Document a recently solved problem or durable project vocabulary in docs/solutions/ or CONCEPTS.md. Use when capturing a learning after work.

ce-pov

by EveryInc

"Give a decisive, project-grounded point of view in the subject's own shape: a graded verdict on an external-adoption question, a holistic take on a document, or a position on a user-supplied approach set. Use for a solo POV, a mid-session second opinion, a named-peer cross-check, any request to consult other models or reconcile their opinions, an oracle panel, or a correction-cost-gated proactive cross-check offer. Not for findings review (use ce-doc-review), neutral explainers, or generating options (use ce-ideate or ce-brainstorm)."

mechsuit

by gravermistakes

figma-to-html-pixel-perfect

by fieldqqq

Convert Figma designs, screenshots, design specifications, and design-system references into production-ready HTML and CSS with maximum visual fidelity. Use this skill whenever the user asks to recreate, implement, audit, improve, or make a website match a Figma design. The skill must distinguish exact design implementation from optional UX/UI enhancements, ask permission before making visible design changes, and verify responsiveness, accessibility, animations, and visual consistency.

task-graph

by igorrendulic

Turn approved implementation plans into plan-isolated task files and a conservative executable DAG in .agent/<plan-slug>/.

ce-optimize

by EveryInc

"Run metric-driven optimization loops. Use when improving measurable outcomes such as search relevance, clustering quality, build performance, prompt quality, or scored behavior through experiments."

ce-debug

by EveryInc

'Diagnosis loop for bugs and failing behavior. Use for errors, stack traces, regressions, failed tests, issue-tracker bugs, stuck investigations after failed fixes, or asks to debug/fix a bug.'

jarvis-ui

by nuwud

JARVIS-style HUD web interface for OpenClaw agents. Interactive Three.js orb with agent state visualization, real-time chat via Gateway WebSocket, audio spectrum analyzer, system monitor, and TTS. Use when you want a visual dashboard for your OpenClaw agent.

learnmap-skill

by lwbscu

Map-first personalized learning skill for Claude Code/Codex. Use when the user wants to learn an unfamiliar domain, turn articles/docs/repos into a curriculum, choose HTML count plus compact/standard/high-quality courseware through native popup calibration, choose batch or multi-turn delivery, receive interactive HTML teaching with portable underlines and image-rich notes, create concept maps, run mastery quizzes and review links, optionally create visual explainers or use a named mentor lens, or use an agent tutor that adapts to stated background and mistakes.

ce-simplify-code

by EveryInc

"Simplify recently changed code for clarity, reuse, quality, and efficiency while preserving behavior. Use for tidy/refactor passes; use ce-debug for bugs."

ibook-builder

by Ibook000

ibook 的工程 builder 视角与执行协议。基于当前仓库中原始人格文档、GitHub 公开资料与项目、 简历履历、以及 Notion 中近期学习笔记蒸馏而成。用途：当用户提到「用 ibook 的方式」 「按 ibook 的思路」「切到 ibook 模式」「别空谈，直接做成能跑的系统」时使用。 适合 AI 应用开发、Python 后端、Agent / MCP / Skill 设计、自动化工具、交易系统、 Web 后台、AIoT 与部署落地场景。这个版本面向公开仓库，既给人看，也给模型用。

vera-language

by aallan

Write programs in the Vera programming language. Use when asked to write, edit, debug, or review Vera code (.vera files). Vera is a statically typed, purely functional language with algebraic effects, mandatory contracts, and typed slot references (@T.n) instead of variable names.