Flowise CVE-2026-41275

Exploitation requires the attacker to occupy a man-in-the-middle position on the same network segment as the victim (e.g., shared public Wi-Fi) AND the victim to actively initiate a password reset and open/transmit the resulting http:// link over that untrusted network during the token's validity window. Limiting factors: no attack is possible without a live network-interception vantage point; if the victim's traffic path is otherwise encrypted (VPN) or the reset is performed on a trusted network, interception fails; the token is single-use and time-limited, so the attacker must capture and replay it before the victim completes their own reset. This is not exploitable against a victim who never triggers a reset, and it is not a remote unauthenticated attack against a default configuration.

The signals are consistent and point to a real but conditional risk rather than a mass-exploitation event. The vendor CVSS 4.0 base score is 7.5, but its own metrics acknowledge the friction: AC:H (high complexity) and AT:P (a specific attack requirement - the attacker must hold a man-in-the-middle position) with UI:A (victim must actively perform a reset and open the link over an untrusted network). SSVC aligns with this: Exploitation=PoC (not active), Automatable=No (each attack requires a live MITM opportunity against a specific user), yet Technical Impact=Total (full account takeover). EPSS of 0.03% and the absence of any KEV listing reinforce that this is not a wormable or widely exploited issue. One nuance to flag: the 'Authentication Bypass' tag overstates the flaw - this is credential/token interception requiring a privileged network position, not a direct auth bypass. Net: patch it as good hygiene, but it is not a drop-everything emergency for most deployments.

A victim on public airport Wi-Fi requests a password reset for their Flowise account and receives an email whose link uses http://. An attacker on the same network passively sniffs the traffic (or performs an active MITM), captures the reset token as it is transmitted in cleartext, and replays it to set a new password and take over the account. A public proof-of-concept demonstrating the insecure HTTP link is documented in HackerOne report #1888915.