<?xml version='1.0' encoding='UTF-8'?>
<?xml-stylesheet href="/static/style.xsl" type="text/xsl"?>
<rss xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/" version="2.0">
  <channel>
    <title>Most recent entries from all</title>
    <link>https://cve.radiocsirt.org</link>
    <description>Contains only the most 10 recent entries.</description>
    <docs>http://www.rssboard.org/rss-specification</docs>
    <generator>python-feedgen</generator>
    <language>en</language>
    <lastBuildDate>Wed, 07 Oct 2026 07:10:14 +0000</lastBuildDate>
    <item>
      <title>EUVD-2026-339216</title>
      <link>https://cve.radiocsirt.org/vuln/euvd-2026-339216</link>
      <description>EUVD-2026-339216</description>
      <guid isPermaLink="false">https://cve.radiocsirt.org/vuln/euvd-2026-339216</guid>
    </item>
    <item>
      <title>fkie_cve-2026-54051</title>
      <link>https://cve.radiocsirt.org/vuln/fkie_cve-2026-54051</link>
      <description>&lt;p&gt;Network-AI is a TypeScript/Node.js multi-agent orchestrator. Prior to version 5.9.1, the agent sandbox gates shell commands behind an allowlist (`SandboxPolicy.isCommandAllowed`), which THREAT_MODEL.md calls the main control against a compromised agent (Adversary 3.2). The allowlist glob-matches the whole command string, but `ShellExecutor` runs that string through `/bin/sh -c`. So any wildcard allow such as `git *`, `npm *` or `node *` also matches `git status; &amp;lt;anything&amp;gt;`, and a scoped command becomes arbitrary execution. The issue is fixed in v5.9.1. `ShellExecutor` now executes via `spawn(file, args, { shell: false })` using a quote-aware parsed argv, so no shell is invoked. `SandboxPolicy.isCommandAllowed` and the new `SandboxPolicy.tokenizeCommand` reject any unquoted shell metacharacter (`; &amp;amp; | $ ` ` ` ( ) &amp;lt; &amp;gt; { }` newline) or unterminated quote before the allowlist glob match; quoted metacharacters are preserved as literal argument data. Users should upgrade to `network-ai@5.9.1` or later. As defense in depth, avoid broad wildcard allowlist entries such as `node *` / `npm *` which are direct code execution by design.&lt;/p&gt;</description>
      <content:encoded>&lt;p&gt;Network-AI is a TypeScript/Node.js multi-agent orchestrator. Prior to version 5.9.1, the agent sandbox gates shell commands behind an allowlist (`SandboxPolicy.isCommandAllowed`), which THREAT_MODEL.md calls the main control against a compromised agent (Adversary 3.2). The allowlist glob-matches the whole command string, but `ShellExecutor` runs that string through `/bin/sh -c`. So any wildcard allow such as `git *`, `npm *` or `node *` also matches `git status; &amp;lt;anything&amp;gt;`, and a scoped command becomes arbitrary execution. The issue is fixed in v5.9.1. `ShellExecutor` now executes via `spawn(file, args, { shell: false })` using a quote-aware parsed argv, so no shell is invoked. `SandboxPolicy.isCommandAllowed` and the new `SandboxPolicy.tokenizeCommand` reject any unquoted shell metacharacter (`; &amp;amp; | $ ` ` ` ( ) &amp;lt; &amp;gt; { }` newline) or unterminated quote before the allowlist glob match; quoted metacharacters are preserved as literal argument data. Users should upgrade to `network-ai@5.9.1` or later. As defense in depth, avoid broad wildcard allowlist entries such as `node *` / `npm *` which are direct code execution by design.&lt;/p&gt;</content:encoded>
      <guid isPermaLink="false">https://cve.radiocsirt.org/vuln/fkie_cve-2026-54051</guid>
    </item>
    <item>
      <title>GHSA-qw6v-5fcf-5666 — Network-AI: Improper Neutralization of Special Elements used in an OS Command</title>
      <link>https://cve.radiocsirt.org/vuln/ghsa-qw6v-5fcf-5666</link>
      <description>&lt;p&gt;&lt;strong&gt;Affected:&lt;/strong&gt; npm: network-ai&lt;/p&gt;
&lt;p&gt;## Summary&lt;/p&gt;
&lt;p&gt;The agent sandbox gates shell commands behind an allowlist (`SandboxPolicy.isCommandAllowed`), which THREAT_MODEL.md calls the main control against a compromised agent (Adversary 3.2). The allowlist glob-matches the whole command string, but `ShellExecutor` runs that string through `/bin/sh -c`. So any wildcard allow such as `git *`, `npm *` or `node *` also matches `git status; &amp;lt;anything&amp;gt;`, and a scoped command becomes arbitrary execution.&lt;/p&gt;
&lt;p&gt;## Root cause&lt;/p&gt;
&lt;p&gt;Matching and execution disagree on what a command is. Lines pinned to `40e42d7` (`lib/agent-runtime.ts` is identical to the v5.8.5 tag).&lt;/p&gt;
&lt;p&gt;1. `isCommandAllowed` matches the full string, with no tokenizing and no metacharacter check:&lt;/p&gt;
&lt;p&gt;https://github.com/Jovancoding/Network-AI/blob/40e42d7a0a966b948953b3c524cf15355d20ef5e/lib/agent-runtime.ts#L248-L260&lt;/p&gt;
&lt;p&gt;2. `globMatch` compiles `*` to `.*` and anchors it, so `git *` becomes `^git .*$` and matches `git status; id`:&lt;/p&gt;
&lt;p&gt;https://github.com/Jovancoding/Network-AI/blob/40e42d7a0a966b948953b3c524cf15355d20ef5e/lib/agent-runtime.ts#L353-L360&lt;/p&gt;
&lt;p&gt;3. `ShellExecutor.execute` only checks `isCommandAllowed`, never `requiresApproval`:&lt;/p&gt;
&lt;p&gt;https://github.com/Jovancoding/Network-AI/blob/40e42d7a0a966b948953b3c524cf15355d20ef5e/lib/agent-runtime.ts#L387-L391&lt;/p&gt;
&lt;p&gt;4. `spawnCommand` runs the approved string via `/bin/sh -c`, so `;`, `|` and `$(...)` are interpreted by the shell:&lt;/p&gt;
&lt;p&gt;https://github.com/Jovancoding/Network-AI/blob/40e42d7a0a966b948953b3c524cf15355d20ef5e/lib/agent-runtime.ts#L427-L43…&lt;/p&gt;</description>
      <content:encoded>&lt;p&gt;&lt;strong&gt;Affected:&lt;/strong&gt; npm: network-ai&lt;/p&gt;
&lt;p&gt;## Summary&lt;/p&gt;
&lt;p&gt;The agent sandbox gates shell commands behind an allowlist (`SandboxPolicy.isCommandAllowed`), which THREAT_MODEL.md calls the main control against a compromised agent (Adversary 3.2). The allowlist glob-matches the whole command string, but `ShellExecutor` runs that string through `/bin/sh -c`. So any wildcard allow such as `git *`, `npm *` or `node *` also matches `git status; &amp;lt;anything&amp;gt;`, and a scoped command becomes arbitrary execution.&lt;/p&gt;
&lt;p&gt;## Root cause&lt;/p&gt;
&lt;p&gt;Matching and execution disagree on what a command is. Lines pinned to `40e42d7` (`lib/agent-runtime.ts` is identical to the v5.8.5 tag).&lt;/p&gt;
&lt;p&gt;1. `isCommandAllowed` matches the full string, with no tokenizing and no metacharacter check:&lt;/p&gt;
&lt;p&gt;https://github.com/Jovancoding/Network-AI/blob/40e42d7a0a966b948953b3c524cf15355d20ef5e/lib/agent-runtime.ts#L248-L260&lt;/p&gt;
&lt;p&gt;2. `globMatch` compiles `*` to `.*` and anchors it, so `git *` becomes `^git .*$` and matches `git status; id`:&lt;/p&gt;
&lt;p&gt;https://github.com/Jovancoding/Network-AI/blob/40e42d7a0a966b948953b3c524cf15355d20ef5e/lib/agent-runtime.ts#L353-L360&lt;/p&gt;
&lt;p&gt;3. `ShellExecutor.execute` only checks `isCommandAllowed`, never `requiresApproval`:&lt;/p&gt;
&lt;p&gt;https://github.com/Jovancoding/Network-AI/blob/40e42d7a0a966b948953b3c524cf15355d20ef5e/lib/agent-runtime.ts#L387-L391&lt;/p&gt;
&lt;p&gt;4. `spawnCommand` runs the approved string via `/bin/sh -c`, so `;`, `|` and `$(...)` are interpreted by the shell:&lt;/p&gt;
&lt;p&gt;https://github.com/Jovancoding/Network-AI/blob/40e42d7a0a966b948953b3c524cf15355d20ef5e/lib/agent-runtime.ts#L427-L43…&lt;/p&gt;</content:encoded>
      <guid isPermaLink="false">https://cve.radiocsirt.org/vuln/ghsa-qw6v-5fcf-5666</guid>
    </item>
  </channel>
</rss>
