<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Oleksandr Prokhorenko</title>
    <description>The latest articles on DEV Community by Oleksandr Prokhorenko (@djminikin).</description>
    <link>https://dev.to/djminikin</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3881987%2Fa850c659-e354-43bd-94cc-1883bb2c36e5.jpg</url>
      <title>DEV Community: Oleksandr Prokhorenko</title>
      <link>https://dev.to/djminikin</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/djminikin"/>
    <language>en</language>
    <item>
      <title>The dependency vulnerability gap that CI/CD can’t fix</title>
      <dc:creator>Oleksandr Prokhorenko</dc:creator>
      <pubDate>Thu, 16 Apr 2026 08:39:31 +0000</pubDate>
      <link>https://dev.to/djminikin/the-dependency-vulnerability-gap-that-cicd-cant-fix-hhn</link>
      <guid>https://dev.to/djminikin/the-dependency-vulnerability-gap-that-cicd-cant-fix-hhn</guid>
      <description>&lt;p&gt;Every project I’ve worked on has the same setup: osv-scanner or Dependabot wired into CI, which fails the build if a known CVE is found. It feels complete. It isn't.&lt;/p&gt;

&lt;p&gt;Here’s the gap: CI runs at push time. CVEs are published continuously.&lt;/p&gt;

&lt;p&gt;If a vulnerability is disclosed in express, serde, or requests the day after your last commit, your CI pipeline won't catch it until your next push. For a team that deploys once a week, that's up to seven days of running known-vulnerable software in production with no alert, no PR, no indication that anything is wrong.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Why Dependabot doesn’t fully solve this&lt;/strong&gt;&lt;br&gt;
Dependabot is genuinely useful, and I’m not dismissing it. But it has real constraints:&lt;/p&gt;

&lt;p&gt;It only works on GitHub-hosted repositories&lt;/p&gt;

&lt;p&gt;It supports around 10 ecosystems&lt;/p&gt;

&lt;p&gt;It watches only the default branch (main/master)&lt;/p&gt;

&lt;p&gt;It opens PRs, which is great for planned maintenance, but not for “you’re vulnerable right now, act fast.”&lt;/p&gt;

&lt;p&gt;If you’re on GitLab, self-hosted Gitea, or running a polyglot monorepo with a mix of Rust, Python, and Go, you’re largely on your own.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;The AI coding tools problem makes this worse&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;AI coding assistants: Claude. Copilot and Cursor introduce dependencies faster than most developers can review them. That’s not a criticism; it’s just the nature of the tooling. But these assistants have a training cutoff. They are unaware of CVEs published after that date.&lt;/p&gt;

&lt;p&gt;Concretely, an AI agent suggests &lt;a href="mailto:somepackage@2.1.0"&gt;somepackage@2.1.0&lt;/a&gt;. At the time, it was suggested that no known vulnerabilities existed. Three weeks later, a CRITICAL CVE is published. Your CI pipeline has no reason to re-run. Nobody gets a notification. You find out when a user files a report.&lt;/p&gt;

&lt;p&gt;The problem isn’t that the AI made a bad suggestion. The problem is that there’s no continuous monitoring layer watching for changes after the fact.&lt;/p&gt;

&lt;p&gt;The right mental model isn’t “scan at push time.” It’s “watch continuously and alert when the threat landscape changes.”&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;The gap, summarised&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F6p6liduogiyt26qgpt90.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F6p6liduogiyt26qgpt90.png" alt=" " width="800" height="160"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;What we built&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;We built &lt;a href="https://oppsy.dev/" rel="noopener noreferrer"&gt;Oppsy&lt;/a&gt; specifically to fill this gap. You upload a lock file, configure a notification channel, and we recheck your dependencies whenever the OSV database is updated. New CVE at 3 am? You’ll know by morning.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://oppsy.dev/" rel="noopener noreferrer"&gt;It’s launching in summer 2026. The waitlist is open now — it’s free to join.&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;If this resonates, I’d be curious whether the notification channel matters to you. Is Slack the obvious choice, or does your team live somewhere else?&lt;/p&gt;

</description>
      <category>automation</category>
      <category>cicd</category>
      <category>devops</category>
      <category>security</category>
    </item>
  </channel>
</rss>
