<?xml version='1.0' encoding='UTF-8'?>
<?xml-stylesheet href="/static/style.xsl" type="text/xsl"?>
<feed xmlns="http://www.w3.org/2005/Atom" xml:lang="en">
  <id>https://vulnerability.circl.lu/sightings/feed</id>
  <title>Most recent sightings.</title>
  <updated>2026-09-28T19:01:19.799279+00:00</updated>
  <author>
    <name>Vulnerability-Lookup</name>
    <email>info@circl.lu</email>
  </author>
  <link href="https://vulnerability.circl.lu" rel="alternate"/>
  <generator uri="https://lkiesow.github.io/python-feedgen" version="1.0.0">python-feedgen</generator>
  <subtitle>Contains only the most 10 recent sightings.</subtitle>
  <entry>
    <id>https://vulnerability.circl.lu/sighting/c816b762-9c08-4c9c-b1fe-3051ad2da119/export</id>
    <title>c816b762-9c08-4c9c-b1fe-3051ad2da119</title>
    <updated>2026-09-28T19:01:19.816109+00:00</updated>
    <author>
      <name>Automation user</name>
      <uri>https://cve.circl.lu/user/automation</uri>
    </author>
    <content>{"uuid": "c816b762-9c08-4c9c-b1fe-3051ad2da119", "vulnerability_lookup_origin": "1a89b78e-f703-45f3-bb86-59eb712668bd", "author": "9f56dd64-161d-43a6-b9c3-555944290a09", "vulnerability": "GHSA-456V-XQ2P-R4CJ", "type": "seen", "source": "https://gist.github.com/alon710/e77a4e3d431108fb441b4cfcda945873", "content": "# GHSA-456V-XQ2P-R4CJ: GHSA-456V-XQ2P-R4CJ: OS Command Injection in code-ollama grep_search Tool\n\n&amp;gt; **CVSS Score:** 7.8\n&amp;gt; **Published:** 2026-09-28\n&amp;gt; **Full Report:** https://cvereports.com/reports/GHSA-456V-XQ2P-R4CJ\n\n## Summary\nAn OS command injection vulnerability in the grep_search tool of the code-ollama package allows remote code execution. This vulnerability is triggered when a local client executes the CLI against a malicious or compromised Ollama server. Due to grep_search being classified as a read-only tool, the CLI executes it automatically in Plan mode without human-in-the-loop validation, leading to zero-interaction local system compromise.\n\n## TL;DR\nA command injection vulnerability in the code-ollama grep_search tool allows compromised LLM backends to run arbitrary local commands without user interaction.\n\n## Exploit Status: POC\n\n## Technical Details\n\n- **CWE ID**: CWE-78\n- **Attack Vector**: Local\n- **CVSS Score**: 7.8\n- **Impact**: Arbitrary OS Command Execution\n- **Exploit Status**: poc\n- **Affected Component**: grep_search tool (grep.ts)\n\n## Affected Systems\n\n- code-ollama npm package\n- **code-ollama**: &amp;lt;= 0.36.0 (Fixed in: `0.36.1`)\n\n## Mitigation\n\n- Upgrade code-ollama to version 0.36.1 or higher\n- Restrain Ollama client connection targets to known, secure endpoints\n- Enforce human-in-the-loop validation for all AI tool operations\n\n**Remediation Steps:**\n1. Verify current installed version of code-ollama using npm list -g code-ollama\n2. Execute npm install -g code-ollama@latest to update to the latest secure release\n3. Confirm OLLAMA_HOST environment variable points to a trusted local or corporate deployment\n\n## References\n\n- [GitHub Security Advisory GHSA-456V-XQ2P-R4CJ](https://github.com/advisories/GHSA-456V-XQ2P-R4CJ)\n- [Internal Security Advisory Mapping](https://github.com/ai-action/code-ollama/security/advisories/GHSA-456v-xq2p-r4cj)\n\n\n---\n*Generated by [CVEReports](https://cvereports.com/reports/GHSA-456V-XQ2P-R4CJ) - Automated Vulnerability Intelligence*", "creation_timestamp": "2026-09-28T14:30:32.000000Z"}</content>
    <link href="https://vulnerability.circl.lu/sighting/c816b762-9c08-4c9c-b1fe-3051ad2da119/export"/>
    <published>2026-09-28T14:30:32+00:00</published>
  </entry>
</feed>
