Guardicore Platform Agent EUVD-2026-100834

Exploitation targets the JWT token validation used during the Guardicore Platform Agent's seeding/enrollment flow. For an unenrolled client, the attacker must convince a user to interact with a malicious link carrying a spoofed JWT seeding token. For an already-seeded client, the attacker must additionally know the specific IDP (identity provider) URL that the client used for enrollment - this is a concrete prerequisite that limits opportunistic attacks. Limiting factors: the attack requires user interaction (per the description, despite the vector's UI:N), is rated high complexity (AC:H), and only affects deployments running a version before 7.4; the IDP-URL knowledge requirement further constrains targeting of established clients.

This is a legitimately serious flaw but not an immediate drop-everything emergency based on current signals. The vendor-supplied CVSS 3.1 vector (CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:C/C:H/I:H/A:H, base 8.5) indicates a network-reachable attack with high attack complexity, low privileges required, a scope change, and full CIA impact - consistent with the described outcome of 'unauthorized reconfiguration and full compromise of the application's functionality.' No EPSS score, no CISA KEV listing, and no public POC were provided, so there is no public exploit identified at time of analysis and no evidence of active exploitation. CONFLICT TO VERIFY: the CVSS vector states UI:N (no user interaction) yet the description explicitly requires 'convincing a user to interact with a malicious link,' which implies UI:R - the vendor's UI:N appears inconsistent with its own description, so treat the 'no user interaction' claim with suspicion. The AC:H rating is corroborated by the description's precondition that attacking an already-seeded client requires the attacker to know the enrollment IDP URL. Net assessment: a real, high-severity priority for organizations running Guardicore Platform Agent below 7.4, but the social-engineering requirement, high attack complexity, and absence of KEV/EPSS/POC signals mean exploitation is likely targeted rather than opportunistic or wormable.

An attacker crafts a malicious link embedding a spoofed JWT seeding token and socially engineers a user into clicking it, causing the Guardicore Platform Agent to accept attacker-controlled parameters and reconfigure itself under the attacker's control. Against an already-enrolled client, the attacker who has learned the client's enrollment IDP URL can target it directly to force reconfiguration. No public exploit code is known at time of analysis, so an attacker would need to develop the token-spoofing technique independently.