A flaw in yt-dlp allows a maliciously crafted URL to trigger command injection when the program processes the "--netrc-cmd" command‑line option (or the equivalent Python API parameter). If an attacker supplies a URL that contains unexpected characters, the downloader will execute those characters as shell commands, giving the attacker full control over the host on which yt‑dlp runs. The vulnerability is classified as CWE‑78, indicating unsafe system call construction. Successful exploitation would compromise confidentiality, integrity, and availability of the affected system.

All users of yt‑dlp versions from 2023‑06‑21 up to, but not including, 2026‑02‑21 are affected. Versions 2026‑02‑21 and later include a fix that validates netrc machine values and rejects invalid input. The product is the yt‑dlp command‑line audio/video downloader distributed under the yt‑dlp_project vendor.

The CVSS score of 8.8 marks this a high‑severity bug. The EPSS score is reported as less than 1 %, suggesting a low probability of exploitation in the wild, and it is not listed in the CISA KEV catalog. The likely attack path requires a user or script to invoke yt‑dlp with the vulnerable option and a malicious URL; this could occur manually or through a web‑hosted redirect that the downloader follows automatically. Because the input is supplied via a URL, the vulnerability is typically exposed only when the user or application trusts external URLs. In absence of a malicious URL, the system remains safe.