<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>its-boris.com</title>
    <link>https://its-boris.com</link>
    <description>AI, agents, and the things we build together. A blog by Boris Barash.</description>
    <language>en-us</language>
    <lastBuildDate>Sat, 18 Apr 2026 11:37:25 GMT</lastBuildDate>
    <atom:link href="https://its-boris.com/feed.xml" rel="self" type="application/rss+xml"/>
    <item>
      <title><![CDATA[AutoResearch: How We Use AI to Improve Our Own Platform]]></title>
      <link>https://its-boris.com/blog/autoresearch-how-we-use-ai-to-improve-our-own-platform</link>
      <guid isPermaLink="true">https://its-boris.com/blog/autoresearch-how-we-use-ai-to-improve-our-own-platform</guid>
      <description><![CDATA[We built an autonomous experimentation pipeline that runs real experiments on our B2B dashboard — the Karpathy Loop with governance, cost tracking, and knowledge accumulation. Here's how it works and what we learned.]]></description>
      <pubDate>Sun, 05 Apr 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>autoresearch</category>
      <category>karpathy-loop</category>
      <category>openclaw</category>
      <category>governance</category>
      <category>self-improving-agents</category>
      <category>devops</category>
    </item>
    <item>
      <title><![CDATA[Live Build — Ship Code to Production in 3 Minutes]]></title>
      <link>https://its-boris.com/blog/live-build-ship-in-3-minutes</link>
      <guid isPermaLink="true">https://its-boris.com/blog/live-build-ship-in-3-minutes</guid>
      <description><![CDATA[Watch Claude Code write real code in real-time, auto-deploy it to production, and verify it with screenshots — all from a single plain-English description. Here's how the 14-stage pipeline works and what the numbers actually look like.]]></description>
      <pubDate>Tue, 31 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>dev-team</category>
      <category>pipeline</category>
      <category>live-build</category>
      <category>deploy</category>
      <category>automation</category>
    </item>
    <item>
      <title><![CDATA[The Week We Stopped Building Features]]></title>
      <link>https://its-boris.com/blog/the-week-we-stopped-building</link>
      <guid isPermaLink="true">https://its-boris.com/blog/the-week-we-stopped-building</guid>
      <description><![CDATA[74 commits, 30 fixes, 4,700 new tests. After a marathon sprint, we spent three days doing nothing but making the platform actually work.]]></description>
      <pubDate>Sat, 28 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>retrospective</category>
      <category>testing</category>
      <category>production</category>
      <category>bugs</category>
      <category>engineering</category>
    </item>
    <item>
      <title><![CDATA[Marathon Session: 4 Sites, 50+ Features, One Weekend]]></title>
      <link>https://its-boris.com/blog/marathon-session-what-we-shipped</link>
      <guid isPermaLink="true">https://its-boris.com/blog/marathon-session-what-we-shipped</guid>
      <description><![CDATA[A full accounting of what got built, broken, and fixed in a single continuous session — from per-org GitHub config to a working AI chat widget.]]></description>
      <pubDate>Wed, 25 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>retrospective</category>
      <category>shipping</category>
      <category>platform</category>
      <category>marathon</category>
    </item>
    <item>
      <title><![CDATA[Why AI Agents Need a Locked Room]]></title>
      <link>https://its-boris.com/blog/why-ai-agents-need-a-locked-room</link>
      <guid isPermaLink="true">https://its-boris.com/blog/why-ai-agents-need-a-locked-room</guid>
      <description><![CDATA[135,000 exposed instances. 63% vulnerable. AI agents are powerful — and that's exactly why they need containers, governance, and an audit trail.]]></description>
      <pubDate>Wed, 25 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>security</category>
      <category>openclaw</category>
      <category>containers</category>
      <category>governance</category>
    </item>
    <item>
      <title><![CDATA[From GitHub Issue to Production Deploy in 3 Minutes — Building an AI Dev Team Pipeline]]></title>
      <link>https://its-boris.com/blog/building-ai-dev-team-pipeline</link>
      <guid isPermaLink="true">https://its-boris.com/blog/building-ai-dev-team-pipeline</guid>
      <description><![CDATA[How we built a fully automated pipeline where a GitHub issue triggers AI decomposition, Docker containers running Claude Code Opus write the code, automated review catches bugs, and a configurable deploy pipeline pushes to production. Real numbers: PR #1235 in 179 seconds, $0.22 total cost.]]></description>
      <pubDate>Tue, 24 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>dev-team</category>
      <category>pipeline</category>
      <category>claude-code</category>
      <category>automation</category>
      <category>deploy</category>
      <category>ci-cd</category>
      <category>docker</category>
    </item>
    <item>
      <title><![CDATA[Week in Review: The AI Dev Team Goes Multi-Repo]]></title>
      <link>https://its-boris.com/blog/week-in-review-ai-dev-team-goes-multi-repo</link>
      <guid isPermaLink="true">https://its-boris.com/blog/week-in-review-ai-dev-team-goes-multi-repo</guid>
      <description><![CDATA[An honest engineering retrospective covering March 19-24, 2026 — 167 commits across 3 repos, 14 merged PRs on a brand-new project, per-org GitHub config, a configurable deploy pipeline, 12 pluggable worker engines, and every bug that bit us along the way. LLM cost: $0.]]></description>
      <pubDate>Tue, 24 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>retrospective</category>
      <category>dev-team</category>
      <category>multi-repo</category>
      <category>pipeline</category>
      <category>claude-code</category>
      <category>automation</category>
      <category>marginmandy</category>
      <category>lessons-learned</category>
    </item>
    <item>
      <title><![CDATA[Dogfooding: How We Use Our Own Dashboard to Ship Better]]></title>
      <link>https://its-boris.com/blog/dogfooding-our-own-governance-dashboard</link>
      <guid isPermaLink="true">https://its-boris.com/blog/dogfooding-our-own-governance-dashboard</guid>
      <description><![CDATA[We build a governance platform for AI agents. We also run AI agents. Here's how using our own dashboard — cost attribution, error tracking, evaluation framework — makes our development process concretely better.]]></description>
      <pubDate>Thu, 19 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>dogfooding</category>
      <category>dashboard</category>
      <category>governance</category>
      <category>cost-tracking</category>
      <category>devops</category>
      <category>continuous-improvement</category>
    </item>
    <item>
      <title><![CDATA[Sprint Retro: What Claude Code and I Actually Shipped (And What Broke)]]></title>
      <link>https://its-boris.com/blog/sprint-retro-what-we-shipped-and-what-broke</link>
      <guid isPermaLink="true">https://its-boris.com/blog/sprint-retro-what-we-shipped-and-what-broke</guid>
      <description><![CDATA[A real engineering retrospective on 10 days of human+AI pair programming — 319 commits, 3 production bugs, 1 billing error that overcharged 100x, and what we're doing differently next sprint.]]></description>
      <pubDate>Thu, 19 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>retrospective</category>
      <category>claude-code</category>
      <category>engineering</category>
      <category>ai-pair-programming</category>
      <category>lessons-learned</category>
    </item>
    <item>
      <title><![CDATA[Why We Switched from MiniMax to StepFun (And Found a 100x Billing Bug)]]></title>
      <link>https://its-boris.com/blog/why-we-switched-from-minimax-to-stepfun</link>
      <guid isPermaLink="true">https://its-boris.com/blog/why-we-switched-from-minimax-to-stepfun</guid>
      <description><![CDATA[A model swap that should have been a one-line change turned into a 46-file investigation, uncovered a critical billing bug, and taught us why model governance needs a single source of truth.]]></description>
      <pubDate>Thu, 19 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>models</category>
      <category>cost-tracking</category>
      <category>governance</category>
      <category>stepfun</category>
      <category>debugging</category>
    </item>
    <item>
      <title><![CDATA[How I Built an AI Gateway for OpenClaw]]></title>
      <link>https://its-boris.com/blog/building-ai-gateway</link>
      <guid isPermaLink="true">https://its-boris.com/blog/building-ai-gateway</guid>
      <description><![CDATA[A technical deep-dive on building a governance gateway that proxies 51 LLM providers, enforces cost caps, scans for PII, and tracks every token — all without changing a single line of application code.]]></description>
      <pubDate>Tue, 17 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>gateway</category>
      <category>architecture</category>
      <category>governance</category>
      <category>curate-me</category>
      <category>openclaw</category>
    </item>
    <item>
      <title><![CDATA[How We Cut AI Agent Costs by 85%]]></title>
      <link>https://its-boris.com/blog/cutting-ai-costs-85-percent</link>
      <guid isPermaLink="true">https://its-boris.com/blog/cutting-ai-costs-85-percent</guid>
      <description><![CDATA[A cost optimization case study: swapping models, adding semantic caching, and migrating to DeepSeek R1 took our multi-agent pipeline from $0.12 per analysis to under $0.02 — with no quality loss.]]></description>
      <pubDate>Tue, 17 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>costs</category>
      <category>optimization</category>
      <category>agents</category>
      <category>curate-me</category>
      <category>case-study</category>
    </item>
    <item>
      <title><![CDATA[Agent Identity: The Enterprise Blind Spot]]></title>
      <link>https://its-boris.com/blog/agent-identity-the-enterprise-blind-spot</link>
      <guid isPermaLink="true">https://its-boris.com/blog/agent-identity-the-enterprise-blind-spot</guid>
      <description><![CDATA[Only 21.9% of organizations treat AI agents as independent identities. The rest have a compliance gap that regulators will notice.]]></description>
      <pubDate>Fri, 13 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>security</category>
      <category>agents</category>
      <category>identity</category>
      <category>governance</category>
      <category>enterprise</category>
    </item>
    <item>
      <title><![CDATA[AGENTS.md, MCP, and A2A: The Three Standards Every Agent Team Needs]]></title>
      <link>https://its-boris.com/blog/agents-md-mcp-and-a2a-the-three-standards</link>
      <guid isPermaLink="true">https://its-boris.com/blog/agents-md-mcp-and-a2a-the-three-standards</guid>
      <description><![CDATA[The protocol landscape for AI agents has crystallized. Here's what each standard does, how they compose together, and what your team should adopt now.]]></description>
      <pubDate>Fri, 13 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>agents</category>
      <category>mcp</category>
      <category>a2a</category>
      <category>standards</category>
      <category>openclaw</category>
    </item>
    <item>
      <title><![CDATA[From Single Agent to Fleet: What Changes]]></title>
      <link>https://its-boris.com/blog/from-single-agent-to-fleet</link>
      <guid isPermaLink="true">https://its-boris.com/blog/from-single-agent-to-fleet</guid>
      <description><![CDATA[Moving from one AI agent to nine required rethinking coordination, cost control, and observability. Here's what we learned.]]></description>
      <pubDate>Fri, 13 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>agents</category>
      <category>architecture</category>
      <category>multi-agent</category>
      <category>curate-me</category>
    </item>
    <item>
      <title><![CDATA[Human-in-the-Loop Patterns for AI Agents]]></title>
      <link>https://its-boris.com/blog/human-in-the-loop-patterns-for-ai-agents</link>
      <guid isPermaLink="true">https://its-boris.com/blog/human-in-the-loop-patterns-for-ai-agents</guid>
      <description><![CDATA[Five human approval patterns we use — from simple approve/reject to iterative feedback loops. How to keep humans in control without slowing agents down.]]></description>
      <pubDate>Fri, 13 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>agents</category>
      <category>hitl</category>
      <category>governance</category>
      <category>patterns</category>
    </item>
    <item>
      <title><![CDATA[Perplexity Computer and the Thin Client Pattern]]></title>
      <link>https://its-boris.com/blog/perplexity-computer-and-the-thin-client-pattern</link>
      <guid isPermaLink="true">https://its-boris.com/blog/perplexity-computer-and-the-thin-client-pattern</guid>
      <description><![CDATA[Perplexity's multi-model agent orchestrator validates the thin client approach we've been building. Here's what we learned — and what we shipped.]]></description>
      <pubDate>Fri, 13 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>agents</category>
      <category>architecture</category>
      <category>thin-client</category>
      <category>perplexity</category>
    </item>
    <item>
      <title><![CDATA[The Agent Observability Gap]]></title>
      <link>https://its-boris.com/blog/the-agent-observability-gap</link>
      <guid isPermaLink="true">https://its-boris.com/blog/the-agent-observability-gap</guid>
      <description><![CDATA[89% of organizations have implemented agent observability. Most still can't trace failures through multi-step workflows. Here's why — and what to do about it.]]></description>
      <pubDate>Fri, 13 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>observability</category>
      <category>agents</category>
      <category>monitoring</category>
      <category>governance</category>
    </item>
    <item>
      <title><![CDATA[The SOUL.md Pattern: Giving Agents Personality]]></title>
      <link>https://its-boris.com/blog/the-soul-md-pattern</link>
      <guid isPermaLink="true">https://its-boris.com/blog/the-soul-md-pattern</guid>
      <description><![CDATA[Every agent on this blog has a SOUL.md file that defines its personality, constraints, and communication style. Here's why this matters more than you'd think.]]></description>
      <pubDate>Fri, 13 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>agents</category>
      <category>openclaw</category>
      <category>patterns</category>
      <category>ai</category>
    </item>
    <item>
      <title><![CDATA[Cost Tracking Across an Agent Fleet]]></title>
      <link>https://its-boris.com/blog/cost-tracking-across-an-agent-fleet</link>
      <guid isPermaLink="true">https://its-boris.com/blog/cost-tracking-across-an-agent-fleet</guid>
      <description><![CDATA[Running 10 AI agents daily gets expensive fast. Here's how we track costs per agent and per model — and how budget caps prevent runaway spending.]]></description>
      <pubDate>Thu, 12 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>costs</category>
      <category>agents</category>
      <category>governance</category>
      <category>curate-me</category>
    </item>
    <item>
      <title><![CDATA[Iterative Refinement: How AI Reviews AI]]></title>
      <link>https://its-boris.com/blog/iterative-refinement-how-ai-reviews-ai</link>
      <guid isPermaLink="true">https://its-boris.com/blog/iterative-refinement-how-ai-reviews-ai</guid>
      <description><![CDATA[How the blog uses a quality loop where one model writes and another reviews — with configurable thresholds, max iterations, and score tracking.]]></description>
      <pubDate>Thu, 12 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>ai</category>
      <category>refinement</category>
      <category>agents</category>
      <category>quality</category>
    </item>
    <item>
      <title><![CDATA[Webhook-Driven Agent Architecture]]></title>
      <link>https://its-boris.com/blog/webhook-driven-agent-architecture</link>
      <guid isPermaLink="true">https://its-boris.com/blog/webhook-driven-agent-architecture</guid>
      <description><![CDATA[Why we chose webhooks over WebSockets or polling for agent communication — and how a single endpoint handles 15 different event types.]]></description>
      <pubDate>Thu, 12 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>architecture</category>
      <category>agents</category>
      <category>webhooks</category>
      <category>patterns</category>
    </item>
    <item>
      <title><![CDATA[Agent Governance: Why It Matters]]></title>
      <link>https://its-boris.com/blog/agent-governance-why-it-matters</link>
      <guid isPermaLink="true">https://its-boris.com/blog/agent-governance-why-it-matters</guid>
      <description><![CDATA[AI agents are powerful — but without governance, they're unpredictable, expensive, and impossible to audit. Here's how curate-me.ai solves this.]]></description>
      <pubDate>Wed, 11 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>governance</category>
      <category>ai</category>
      <category>agents</category>
      <category>curate-me</category>
    </item>
    <item>
      <title><![CDATA[How the Agent Fleet Actually Works]]></title>
      <link>https://its-boris.com/blog/how-the-agent-fleet-actually-works</link>
      <guid isPermaLink="true">https://its-boris.com/blog/how-the-agent-fleet-actually-works</guid>
      <description><![CDATA[A deep dive into the 9 OpenClaw agents that run this blog — their profiles, schedules, tools, and how they coordinate through the curate-me.ai gateway.]]></description>
      <pubDate>Wed, 11 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>agents</category>
      <category>openclaw</category>
      <category>architecture</category>
      <category>curate-me</category>
    </item>
    <item>
      <title><![CDATA[Building a Blog With AI Agents]]></title>
      <link>https://its-boris.com/blog/building-a-blog-with-ai-agents</link>
      <guid isPermaLink="true">https://its-boris.com/blog/building-a-blog-with-ai-agents</guid>
      <description><![CDATA[How I rebuilt my blog from scratch using OpenClaw agents governed by curate-me.ai — and why the blog itself is the best reference app for the platform.]]></description>
      <pubDate>Tue, 10 Mar 2026 00:00:00 GMT</pubDate>
      <author>boris@its-boris.com (Boris Barash)</author>
      <category>ai</category>
      <category>agents</category>
      <category>openclaw</category>
      <category>curate-me</category>
    </item>
  </channel>
</rss>