agent-experience-design

Design products that AI agents can effectively use. Covers Agent Experience (AX) four pillars (Access, Context, Tools, Orchestration), agent-readable site standards (llms.txt, AGENTS.md, AgentReady, Lighthouse agentic browsing, MCP Server Cards), mutual legibility (two-way agent-human communication, three counterintuitive HCI findings), the bifurcated web (visible vs invisible, brand-to-bot, AVIO replacing SEO), and agent self-serve implementation (programmatic signup, sandbox provisioning, capability keys, removing human gates, WebMCP). Use when building agent-facing APIs, MCP servers, agent signup flows, agent-readable documentation, pricing for agent buyers, or any product surface where AI agents are users. Use when reviewing or designing agent UX, agent onboarding, agent auth, or agent billing.

跳到安装

仓库: katrinalaszlo/personal-site
最近来源活动: 2026年5月18日 19:16
检测到的 SKILL.md 语言: 英语
星标: 0
分支: 0

安装方式

默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。

用 Codex 或 Claude 帮你安装复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。

直接命令不会经过审查 Prompt；运行前请先检查来源。

npx skills add https://github.com/katrinalaszlo/personal-site --skill agent-experience-design

命令会保持在同一行。复制前请横向滚动并检查完整内容。

想先保存到本地？可下载 SkillsMP 当前能够提供的文件。

检查来源文件

决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。

文件资源管理器

6 个文件

SKILL.md

来源说明 · 只读预览

返回搜索返回安装

katrinalaszlo

Agent experience design, SaaS pricing strategy, and agent-readiness audits. MCP server with notebook search and 5 installable Agent Skills.

2026-08-11

agent-tools

Specific agent platforms and orchestration tools. Covers OpenClaw personal agents (local-first agent framework, personal agent architecture, tool management, memory persistence, multi-agent coordination) and Hermes orchestration (message-based agent coordination, pub/sub patterns, agent-to-agent communication, workflow orchestration). Use when building personal AI agents, setting up agent orchestration infrastructure, or evaluating agent frameworks and coordination patterns.

2026-05-18

ai-foundations

How neural networks work from tokens to predictions. Covers transformer model architecture (embeddings, attention, Q/K/V projections, multi-head attention, RoPE positional encoding, MLPs, residual connections, output head), training loops (forward pass, backward pass, gradients, loss functions, optimization, learning rate schedules, overfitting, regularization), and data pipelines (tokenization, BPE, batching, packing, evaluation metrics). Use when learning or explaining how LLMs work internally, debugging model behavior, or understanding why a model produces certain outputs.

2026-05-18

查看这个仓库的全部 Skills

name	agent-experience-design
description	Design products that AI agents can effectively use. Covers Agent Experience (AX) four pillars (Access, Context, Tools, Orchestration), agent-readable site standards (llms.txt, AGENTS.md, AgentReady, Lighthouse agentic browsing, MCP Server Cards), mutual legibility (two-way agent-human communication, three counterintuitive HCI findings), the bifurcated web (visible vs invisible, brand-to-bot, AVIO replacing SEO), and agent self-serve implementation (programmatic signup, sandbox provisioning, capability keys, removing human gates, WebMCP). Use when building agent-facing APIs, MCP servers, agent signup flows, agent-readable documentation, pricing for agent buyers, or any product surface where AI agents are users. Use when reviewing or designing agent UX, agent onboarding, agent auth, or agent billing.
metadata	{"author":"Katrina Laszlo","version":"1.0.0"}