<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>ATHelper Blog</title>
    <link>https://www.at-helper.com/blog</link>
    <description>Insights on AI-powered testing, automation, and product updates</description>
    <language>en</language>
    <lastBuildDate>Fri, 12 Jun 2026 05:36:05 GMT</lastBuildDate>
    <atom:link href="https://www.at-helper.com/feed" rel="self" type="application/rss+xml"/>
    
    <item>
      <title><![CDATA[Four CVEs in a week, all the same shape: when agents execute LLM-generated code]]></title>
      <link>https://www.at-helper.com/blog/four-cves-in-a-week-all-the-same-shape-when-agents-execute-llm-generated-code</link>
      <guid isPermaLink="true">https://www.at-helper.com/blog/four-cves-in-a-week-all-the-same-shape-when-agents-execute-llm-generated-code</guid>
      <description><![CDATA[Between 2026-05-04 and 2026-05-06, NVD published four CVEs against AI/agent projects that share a single shape: an LLM produces output, the application drops that output into a privileged execution sink without re-validation, and the sink runs it. SQLBot, PPTAgent, Evolver, Dify — different teams, same defect class (OWASP LLM05). The control belongs at the seam between LLM and sink, not on the prompt side.]]></description>
      <pubDate>Thu, 07 May 2026 01:53:16 GMT</pubDate>
      
      <dc:creator><![CDATA[Yang Gao]]></dc:creator>
    </item>
    <item>
      <title><![CDATA[Why Your Agent Eval Suite Is a Security Audit, Not a QA Exercise]]></title>
      <link>https://www.at-helper.com/blog/why-your-agent-eval-suite-is-a-security-audit-not-a-qa-exercise</link>
      <guid isPermaLink="true">https://www.at-helper.com/blog/why-your-agent-eval-suite-is-a-security-audit-not-a-qa-exercise</guid>
      <description><![CDATA[Most engineering teams are building AI agent evaluation suites the way they built QA test suites: a battery of pass/fail checks that gate CI and produce a green badge. That model is structurally wrong for agents. Agent failures don't come from the input distribution your tests cover — they come from the adversarial distribution your tests don't. The correct mental model is the security audit: rotational, adversarial, owned by people whose job is to find what breaks rather than to confirm what works. If your eval suite produces a single number that goes up and to the right, you are not measuring agent reliability. You are measuring how good your agent is at passing your own test set.]]></description>
      <pubDate>Thu, 30 Apr 2026 05:53:35 GMT</pubDate>
      
      <dc:creator><![CDATA[Yang Gao]]></dc:creator>
    </item>
    <item>
      <title><![CDATA[The Wild West of AI Agents: 88% Report Security Incidents — How Testing Can Save You]]></title>
      <link>https://www.at-helper.com/blog/the-wild-west-of-ai-agents-88-report-security-incidents-how-testing-can-save-you</link>
      <guid isPermaLink="true">https://www.at-helper.com/blog/the-wild-west-of-ai-agents-88-report-security-incidents-how-testing-can-save-you</guid>
      <description><![CDATA[88% of organizations reported AI agent security incidents in 2026, yet only 14.4% deploy with full security approval. We explore the OWASP Agentic AI Top 10, Microsoft Agent Governance Toolkit, NIST standards, and why AI-powered testing is the missing security layer.]]></description>
      <pubDate>Wed, 08 Apr 2026 00:25:43 GMT</pubDate>
      
      <dc:creator><![CDATA[Yang Gao]]></dc:creator>
    </item>
    <item>
      <title><![CDATA[Tech News Meets AI Testing: How This Week's Biggest Stories Shape the Future of QA]]></title>
      <link>https://www.at-helper.com/blog/tech-news-meets-ai-testing-how-this-weeks-biggest-stories-shape-the-future-of-qa</link>
      <guid isPermaLink="true">https://www.at-helper.com/blog/tech-news-meets-ai-testing-how-this-weeks-biggest-stories-shape-the-future-of-qa</guid>
      <description><![CDATA[From $297B in AI funding to GPU security vulnerabilities to California's AI regulations — we break down the biggest tech stories of April 2026 and explore what each means for the future of AI-powered software testing and agentic QA.]]></description>
      <pubDate>Wed, 08 Apr 2026 00:10:37 GMT</pubDate>
      
      <dc:creator><![CDATA[Yang Gao]]></dc:creator>
    </item>
    <item>
      <title><![CDATA[How AI Browser Agents Find Bugs That Manual Testing Misses]]></title>
      <link>https://www.at-helper.com/blog/ai-browser-agents-find-bugs-manual-testing-misses</link>
      <guid isPermaLink="true">https://www.at-helper.com/blog/ai-browser-agents-find-bugs-manual-testing-misses</guid>
      <description><![CDATA[Manual testing catches the obvious issues. AI browser agents catch everything else — from subtle race conditions to broken flows across hundreds of viewport sizes. Here is how they work, and why teams adopting them are finding 3x more bugs before production.]]></description>
      <pubDate>Wed, 01 Apr 2026 19:09:25 GMT</pubDate>
      <category>Tutorials</category>
      <dc:creator><![CDATA[Yang Gao]]></dc:creator>
    </item>
    <item>
      <title><![CDATA[2026 Is the Year QA Goes Agentic: What It Means for Your Testing Strategy]]></title>
      <link>https://www.at-helper.com/blog/2026-agentic-qa-testing-strategy</link>
      <guid isPermaLink="true">https://www.at-helper.com/blog/2026-agentic-qa-testing-strategy</guid>
      <description><![CDATA[AI agents are reshaping software testing from scripted automation to autonomous, self-healing workflows. Here is what the agentic QA revolution means for engineering teams — and how to stay ahead.]]></description>
      <pubDate>Wed, 01 Apr 2026 18:28:46 GMT</pubDate>
      <category>Testing Best Practices</category>
      <dc:creator><![CDATA[Yang Gao]]></dc:creator>
    </item>
  </channel>
</rss>