Topics
Use Cases & Examples
What kinds of research sites you can build with AgentPages: project docs, news trackers, competitive intelligence, domain knowledge bases, and personal learning journals.
Model Context Protocol (MCP)
The open standard connecting AI agents to tools and services. How gh-aw uses MCP to give agents access to GitHub, Tavily, Notion, and hundreds of other integrations.
Tavily API
The AI-native search engine powering AgentPages research. Learn about Search, Extract, QnA, Crawl, and how to stay within the free tier's 1,000 credits/month.
Astro Framework
Why AgentPages uses Astro: the islands architecture, zero-JS-by-default, file-based routing, and how the AI agent updates the site on each research run.
How AgentPages Works
The core loop: how the AI agent researches topics, maintains a knowledge base, builds an Astro static site, and publishes via GitHub Pages — all automatically.
GitHub Agentic Workflows
A deep dive into gh-aw, the platform powering AgentPages. Natural-language markdown workflows, supported AI engines (Claude, Copilot, Codex, Gemini), MCP tools, and the security architecture.
Getting Started
Step-by-step setup guide: fork the repo, install gh-aw, configure API secrets, define your research focus, and trigger your first agent run.
AI Engine Comparison
Claude vs Copilot vs Codex vs Gemini for research agents: strengths, pricing, configuration options, and how to choose the right engine for your AgentPages instance.
Continuous AI Patterns
The emerging paradigm of systematic, automated AI applied to software collaboration. Real patterns and metrics from GitHub Next's 19-workflow agent factory.