<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>execute42 · Engineering Notes</title><description>The portfolio of yuyukosama2004: full-stack AI products, RAG systems, agent tooling, and notes on how they are built and tested.</description><link>https://www.execute42.top/</link><language>en</language><item><title>Building an Evidence-backed Web Research MCP</title><link>https://www.execute42.top/en/blog/building-evidence-backed-research-mcp/</link><guid isPermaLink="true">https://www.execute42.top/en/blog/building-evidence-backed-research-mcp/</guid><description>Moving from search results to reviewable conclusions requires source snapshots, claim links, conflict handling, and versioned artifacts—not just a browser tool.</description><pubDate>Tue, 14 Jul 2026 00:00:00 GMT</pubDate><category>MCP</category><category>Web Research</category><category>Verifiable AI</category></item><item><title>Why Long-form AI Writing Needs a Canon Boundary</title><link>https://www.execute42.top/en/blog/why-ai-writing-needs-canon-boundaries/</link><guid isPermaLink="true">https://www.execute42.top/en/blog/why-ai-writing-needs-canon-boundaries/</guid><description>Models can generate plot ideas and prose quickly, but long-form work stays coherent only when suggestions, drafts, and accepted facts remain separate.</description><pubDate>Tue, 14 Jul 2026 00:00:00 GMT</pubDate><category>AI Writing</category><category>Human-in-the-loop</category><category>Architecture</category></item></channel></rss>