Markdown Preview Enhanced EUVD-2026-101583

Exploitation requires that the victim open attacker-controlled Markdown/repository content in a VS Code workspace with Markdown Preview Enhanced (< 0.8.30) active, AND that the attacker already has JavaScript executing inside the preview webview - this is not inherent to this CVE and must be obtained via a separate flaw such as the .crossnote/head.html script injection (GHSA-mcwg-4j78-qwv3) or the blind command-dispatch weakness fixed in the same release. Given that chained context, the attacker can call _crossnote.updateMarkdown with an arbitrary target URI because pre-0.8.30 code did not verify the URI matched the preview's own source file nor enforce a markdown file extension. Limiting factors: local attack vector (malicious content must be opened by the user), the mandatory second execution-path vulnerability, and that writes are bounded by the VS Code user's OS filesystem privileges.

This is a moderate, conditional risk rather than a drop-everything priority. The vendor CVSS 4.0 vector (AV:L/AC:L/AT:P/PR:N/UI:P, VC:N/VI:H/VA:H) scores 5.8 (Medium), and the signals are internally consistent with the description: AV:L and AT:P reflect that exploitation is not a standalone remote attack but requires a local/attacker-supplied context plus a separate precondition. Crucially, the description states this file-write is NOT independently exploitable - it requires a separate webview JavaScript execution path (the independently tracked head.html script injection and blind command-dispatch weaknesses, fixed in the same 0.8.30 release). No EPSS score, CISA KEV listing, or public POC was provided in the input, so there is no evidence of active or at-scale exploitation; the 'RCE' tag reflects the realistic end impact (writing startup/config files) rather than a direct single-shot remote code execution. The dependency on chaining, combined with AV:L and required user interaction (opening a malicious repository/preview), substantially limits real-world risk despite the high integrity/availability impact. Note a minor signal nuance: the input tag is 'RCE' while CWE-73 and the fix describe an arbitrary file-write primitive - code execution is a downstream consequence, not the primitive itself.

A developer opens a malicious or compromised repository in VS Code that is rendered by Markdown Preview Enhanced. Using a chained webview JavaScript execution flaw (e.g., head.html script injection), attacker-controlled code in the preview webview invokes _crossnote.updateMarkdown through the extension message bridge to write an arbitrary file - such as ~/.bashrc, authorized_keys, or a build/startup script - which executes with the user's OS privileges. No public exploit code was identified at time of analysis, and exploitation depends on the separate, now-patched execution path.