<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/"><channel><title>Agentic Engineering in Action (agentprose demo)</title><link>https://vibeprogrammer.github.io/agentprose-demo/</link><description>Demo blog for agentprose: real issues of Maxim Salnikov's LinkedIn newsletter Agentic Engineering in Action, served to AI agents through WebMCP.</description><language>en-US</language><atom:link href="https://vibeprogrammer.github.io/agentprose-demo/index.xml" rel="self" type="application/rss+xml"/><lastBuildDate>Wed, 30 Sep 2026 05:15:04 +0000</lastBuildDate><item><title>Where Is the Gate? A Practical Look at AI Engineering Systems</title><link>https://vibeprogrammer.github.io/agentprose-demo/posts/where-gate-practical-look-ai-engineering-systems/</link><guid>https://vibeprogrammer.github.io/agentprose-demo/posts/where-gate-practical-look-ai-engineering-systems/</guid><pubDate>Wed, 30 Sep 2026 05:15:04 +0000</pubDate><dc:creator>Maxim Salnikov</dc:creator><category>engineering-systems</category><category>ci</category><category>coding-agents</category><description>By an AI engineering system, I mean how a team uses agents to take an idea all the way to a reliable change: the context agents can use, the actions they can take, the tests and approvals a change… (Originally published on LinkedIn: https://www.linkedin.com/pulse/where-gate-practical-look-ai-engineering-systems-maxim-salnikov-g1snc)</description></item><item><title>One corpus, many harnesses: what it took to keep our AI consistent</title><link>https://vibeprogrammer.github.io/agentprose-demo/posts/one-corpus-many-harnesses-what-took-keep-our-ai/</link><guid>https://vibeprogrammer.github.io/agentprose-demo/posts/one-corpus-many-harnesses-what-took-keep-our-ai/</guid><pubDate>Mon, 28 Sep 2026 05:15:20 +0000</pubDate><dc:creator>Einar Ingebrigtsen</dc:creator><category>harness</category><category>context-engineering</category><category>guest-post</category><description>Guest post: This edition is written by Einar Ingebrigtsen, sharing his own perspective and experience. (Originally published on LinkedIn: https://www.linkedin.com/pulse/one-corpus-many-harnesses-what-took-keep-our-ai-maxim-salnikov-vs6pe)</description></item><item><title>What I learned building 6 apps and solving 10 related engineering problems for the OpenAI WebMCP Challenge</title><link>https://vibeprogrammer.github.io/agentprose-demo/posts/what-i-learned-building-6-apps-solving-10-related-openai/</link><guid>https://vibeprogrammer.github.io/agentprose-demo/posts/what-i-learned-building-6-apps-solving-10-related-openai/</guid><pubDate>Fri, 25 Sep 2026 05:30:17 +0000</pubDate><dc:creator>Maxim Salnikov</dc:creator><category>webmcp</category><category>openai</category><category>prototyping</category><description>I built six working app prototypes in a short time for the OpenAI WebMCP Challenge. (Originally published on LinkedIn: https://www.linkedin.com/pulse/what-i-learned-building-6-apps-solving-10-related-openai-salnikov-r0f9e)</description></item><item><title>GitHub Agentic Workflows: The Five-Minute Version + Free Book</title><link>https://vibeprogrammer.github.io/agentprose-demo/posts/github-agentic-workflows-five-minute-version-free-book/</link><guid>https://vibeprogrammer.github.io/agentprose-demo/posts/github-agentic-workflows-five-minute-version-free-book/</guid><pubDate>Mon, 21 Sep 2026 05:45:47 +0000</pubDate><dc:creator>Maxim Salnikov</dc:creator><category>github-agentic-workflows</category><category>books</category><description>What gh-aw is, where it fits, and what to expect from my new interactive book Note: this article is not a book itself (see below) but a quick introduction to GitHub Agentic Workflows so you could… (Originally published on LinkedIn: https://www.linkedin.com/pulse/github-agentic-workflows-five-minute-version-free-book-maxim-salnikov-6nhre)</description></item><item><title>Your Agent’s Context Has a Supply Chain. Manage It.</title><link>https://vibeprogrammer.github.io/agentprose-demo/posts/your-agents-context-has-supply-chain-manage/</link><guid>https://vibeprogrammer.github.io/agentprose-demo/posts/your-agents-context-has-supply-chain-manage/</guid><pubDate>Tue, 08 Sep 2026 20:14:00 +0000</pubDate><dc:creator>Maxim Salnikov</dc:creator><category>agent-package-manager</category><category>context-engineering</category><category>ci</category><description>Building a supply chain for agent context with Microsoft Agent Package Manager An AI coding agent’s working environment starts taking shape before we send the first prompt. (Originally published on LinkedIn: https://www.linkedin.com/pulse/your-agents-context-has-supply-chain-manage-maxim-salnikov-n2e7e)</description></item><item><title>Giving AI agents a real computer</title><link>https://vibeprogrammer.github.io/agentprose-demo/posts/giving-ai-agents-real-computer/</link><guid>https://vibeprogrammer.github.io/agentprose-demo/posts/giving-ai-agents-real-computer/</guid><pubDate>Thu, 23 Jul 2026 06:32:10 +0000</pubDate><dc:creator>Maxim Salnikov</dc:creator><category>microsoft-agent-framework</category><category>sandboxes</category><category>github-copilot</category><description>How I built Sandcastle on the new GitHub Copilot provider for Microsoft Agent Framework — and what the feature actually unlocks ▶️ Live demo: https://sandcastle.isainative. (Originally published on LinkedIn: https://www.linkedin.com/pulse/giving-ai-agents-real-computer-maxim-salnikov-zexxe)</description></item><item><title>The Best AI Coding Model Is a Moving Target. Build a Selection System Instead</title><link>https://vibeprogrammer.github.io/agentprose-demo/posts/best-ai-coding-model-moving-target-build-selection-system/</link><guid>https://vibeprogrammer.github.io/agentprose-demo/posts/best-ai-coding-model-moving-target-build-selection-system/</guid><pubDate>Wed, 15 Jul 2026 05:15:06 +0000</pubDate><dc:creator>Maxim Salnikov</dc:creator><category>evaluation</category><category>models</category><category>coding-agents</category><description>Choosing an AI coding model from a leaderboard is like choosing a route from a screenshot of yesterday's traffic. The route may have been fastest when it was measured. (Originally published on LinkedIn: https://www.linkedin.com/pulse/best-ai-coding-model-moving-target-build-selection-system-salnikov-idvfe)</description></item><item><title>The First 80% Takes Nine Minutes. The Last 20% Takes Two Days.</title><link>https://vibeprogrammer.github.io/agentprose-demo/posts/first-80-takes-nine-minutes-last-20-two-days/</link><guid>https://vibeprogrammer.github.io/agentprose-demo/posts/first-80-takes-nine-minutes-last-20-two-days/</guid><pubDate>Mon, 13 Jul 2026 05:15:03 +0000</pubDate><dc:creator>Maxim Salnikov</dc:creator><category>engineering-systems</category><category>planning</category><category>coding-agents</category><description>Picture a product team at 9:07 on a Tuesday morning. A product manager asks an agent to add saved views to an internal dashboard. (Originally published on LinkedIn: https://www.linkedin.com/pulse/first-80-takes-nine-minutes-last-20-two-days-maxim-salnikov-ingde)</description></item><item><title>You can't tune what you can't see: prompt‑cache observability for GitHub Copilot with OpenTelemetry</title><link>https://vibeprogrammer.github.io/agentprose-demo/posts/you-cant-tune-what-see-promptcache-observability-github/</link><guid>https://vibeprogrammer.github.io/agentprose-demo/posts/you-cant-tune-what-see-promptcache-observability-github/</guid><pubDate>Wed, 08 Jul 2026 12:42:31 +0000</pubDate><dc:creator>Maxim Salnikov</dc:creator><category>github-copilot</category><category>opentelemetry</category><category>observability</category><description>Open VS Code, ask Copilot Chat one question, and if you crack open the debug view you can watch prompt‑cache tokens tick past for that single call. (Originally published on LinkedIn: https://www.linkedin.com/pulse/you-cant-tune-what-see-promptcache-observability-github-salnikov-gstye)</description></item><item><title>Design, but with tests: an engineering approach to UX with Impeccable</title><link>https://vibeprogrammer.github.io/agentprose-demo/posts/design-tests-engineering-approach-ux-impeccable/</link><guid>https://vibeprogrammer.github.io/agentprose-demo/posts/design-tests-engineering-approach-ux-impeccable/</guid><pubDate>Wed, 01 Jul 2026 14:50:55 +0000</pubDate><dc:creator>Maxim Salnikov</dc:creator><category>ux</category><category>testing</category><category>github-copilot</category><description>Impeccable — now built into the GitHub Copilot app — treats design quality the way we treat code quality. This is what that looked like on a real app that was already shipping. (Originally published on LinkedIn: https://www.linkedin.com/pulse/design-tests-engineering-approach-ux-impeccable-maxim-salnikov-f2rle)</description></item><item><title>The model is a commodity. The harness is your advantage. A hands-on experiment with GitHub Copilot app's Canvas</title><link>https://vibeprogrammer.github.io/agentprose-demo/posts/model-commodity-harness-your-advantage-hands-on-github/</link><guid>https://vibeprogrammer.github.io/agentprose-demo/posts/model-commodity-harness-your-advantage-hands-on-github/</guid><pubDate>Tue, 30 Jun 2026 12:29:09 +0000</pubDate><dc:creator>Maxim Salnikov</dc:creator><category>harness</category><category>github-copilot</category><category>canvas</category><description>The teams getting the most out of AI agents are not the ones with the best model. They are the ones building a custom layer on top of it. (Originally published on LinkedIn: https://www.linkedin.com/pulse/model-commodity-harness-your-advantage-hands-on-github-maxim-salnikov-s9mke)</description></item><item><title>Stop Putting a Raw LLM in Your CI Pipeline — Put a Sandbox Around It</title><link>https://vibeprogrammer.github.io/agentprose-demo/posts/stop-putting-raw-llm-your-ci-pipeline-put-sandbox-around/</link><guid>https://vibeprogrammer.github.io/agentprose-demo/posts/stop-putting-raw-llm-your-ci-pipeline-put-sandbox-around/</guid><pubDate>Mon, 29 Jun 2026 14:00:15 +0000</pubDate><dc:creator>Maxim Salnikov</dc:creator><category>github-agentic-workflows</category><category>security</category><category>ci</category><description>There's a seductive idea floating around every engineering org that has discovered GitHub Actions and GitHub Copilot in the same quarter: what if I (Originally published on LinkedIn: https://www.linkedin.com/pulse/stop-putting-raw-llm-your-ci-pipeline-put-sandbox-around-salnikov-qtwme)</description></item></channel></rss>