ucsahinn/codex-enterprise-prompt-architect
Enterprise-grade Codex prompt architect skill with plan-only, approval-gated execution, browser QA, security constraints, and prompt evaluation workflows.
Enterprise-grade Codex prompt architect skill with plan-only, approval-gated execution, browser QA, security constraints, and prompt evaluation workflows.
npx add-skill ucsahinn/codex-enterprise-prompt-architectEnterprise-grade Codex prompt architect skill with plan-only, approval-gated execution, browser QA, security constraints, and prompt evaluation workflows.
Seurat v5 single-cell analysis skill for Claude Code, OpenClaw, and Codex
Long-term memory skill for AI coding agents — self-hosted RAG with semantic search, multimodal documents, and knowledge graph. Works with Claude Code, OpenClaw, Codex CLI.
Claude Code skill for daily AI digest across 5 categories
Autonomous Goal-directed Iteration for Claude Code. Inspired by Karpathy's autoresearch. 9 modes: loop, plan, debug, fix, security, ship, scenario, predict, learn.
Claude Skill，围绕行业、公司或主题做深度研究。采用"先想后搜"方法：先明确研究边界，用 MECE 原则拆解关键子问题，再基于子问题去发现报告定向搜索研报、筛选核心研报，精读后沉淀事实、对比推导，最终形成结构化结论。
Design exploration skill for Claude Code. Research-first: scans forums, competitors, and user pain before proposing anything. Three progressive rounds (literal → creative → wild) produce self-executing specs with embedded implementation instructions.