Art & Design

research,agent-memory,experiment-tracking,agent-skills,skill-md,codex,claude-code,external-memory,ai-memory,hypotheses,experiments

Guide users step-by-step to set up and use OpenClaw AI assistants in Chinese, from chat basics to multi-agent collaboration without coding skills.

Safety-bounded AI agent for disaster triage: reusable Python core, Gemma 4 demo, memory/skill loop, validation gates, audit trace.

Evidence-bounded Codex skill for academic manuscript revision, traceability, figure/table QA, and submission readiness checks.

Monitor Aave liquidation risks and track yields with a high-performance Rust bot that integrates as an OpenClaw skill for DeFi automation.

Claude skills and GitHub Actions workflows for AI-assisted Salesforce development — from issue to pull request, scratch org deployment, and Playwright-based UI validation.

Track Claude Code CLI state with a clean spinner word registry built in vanilla JavaScript and CSS3

Build a tiny Claude Code clone in Python with agent loop, tools, permissions, and context management in ~800 lines

Run Claude Code locally with TUI and Electron desktop support, plus flexible Anthropic/OpenAI-compatible API access and local fixes

Manage persistent session memory and context handoff for Claude Code to maintain continuity between tasks.

Explore reverse engineering specs for Claude Code CLI, with implementation-ready design docs from source map analysis of the npm package

Build high-quality Jamdesk documentation faster with this Claude Code plugin that provides AI with deep knowledge of components, configuration, and CLI commands.

Automate BigQuery tasks with Claude Code. Use this plugin to generate SQL, optimize costs, design schemas, and detect anti-patterns.

Recently updated GitHub repository for pdf to md mineru skill, discovered in the daily agent-skills hunt.

Provide accurate Jetpack Compose code by detecting common errors and improving Android UI practices with a modern agent skill compatible across platforms.

Build a self-hosted health AI agent with open-source OpenClaw skills, private data, and Telegram access on low-bandwidth devices

Portable AI market intelligence scan Agent Skill: fetch signals, save a Markdown recap, render a newspaper-style HTML, export a PDF.

每次 AI 调用都在消耗真实的能源和你的钱包。Better Earth 是一个 Agent Skill，通过引导 AI 批量操作、充分思考后再调用、榨干每次调用的价值，来减少不必要的工具调用——省调用，护地球。

Skill that auto-builds & deploys a Hebrew WhatsApp agent (Kapso + Gemini + Maton). One-click: npx skills add amichaiai/whatsapp-agent-skill

🐕 Self-hosted security-first AI agent · 22 providers · 11 chat channels · WebUI · 3-level memory · task-schedule · automation · skills · MCP

Connect LINE to Claude Code with an MCP server for DMs, group chats, media downloads, and image replies

Templates and guides for structuring a production-grade Claude Code harness. Covers global setup, per-project AGENTS.md, quality gates, memory system, and an invocable Claude Code skill.

Shared Claude Code config for Canibuild-Ops leaders — skills, hooks, settings, org policy CLAUDE.md. One-command bootstrap.

Free, self-contained Claude Code skills (markdown playbooks): PIV loop, CLAUDE.md architecture, commit strategy, prompt template. EN/JA.