Frappe Framework EUVD-2026-102569

Two concrete prerequisites must both hold: (1) the attacker must possess valid non-session credentials for a target user - specifically an API key/secret, an OAuth access token, or a header-based authentication secret (session-cookie logins are NOT affected); and (2) the targeted user account must have a per-user restrict_ip allowlist configured, otherwise there is no network restriction to bypass and the flaw has no effect. The bypass therefore only materializes in deployments that deliberately use restrict_ip as an access control. Limiting factors: this is a privilege-bounded bypass (PR:L) - it grants no more than the stolen credential already allows and does not by itself obtain those credentials; it has no impact on default configurations that do not set restrict_ip; and it is a confidentiality/integrity-partial, availability-none issue.

This is a medium-severity access-control bypass (CVSS 4.0 base 5.3, vendor-supplied vector AV:N/AC:L/AT:N/PR:L/UI:N) with limited technical impact (VC:L/VI:L/VA:N, no scope change). The attack vector is network and complexity is low, but crucially PR:L means the attacker must already hold valid credentials (an API key, OAuth token, or header-auth secret), so this is not a path to initial compromise - it defeats a defense-in-depth network restriction (the per-user restrict_ip allowlist) rather than granting new access. No EPSS score, no CISA KEV listing, and no public exploit were provided in the intelligence, placing this well below the priority of a remote unauthenticated RCE. The SSVC framework is not supplied; inferring from the data, exploitation is not confirmed in the wild, technical impact is partial, and automatable value is low because the bypass only matters once credentials are already compromised. Net assessment: a real but second-order hardening fix most relevant to organizations that deliberately rely on restrict_ip as a compensating control, not a drop-everything emergency. One caveat to verify: the vendor vector assigns VC:L/VI:L, but the actual confidentiality/integrity exposure is entirely determined by the privileges of the stolen credential, so impact could be higher or lower than the base score suggests depending on the account.

An attacker who has obtained a valid Frappe API key or OAuth token for a user whose account is locked down with a restrict_ip allowlist (for example, a leaked key intended to be usable only from a corporate VPN egress IP) authenticates to the REST API from an arbitrary internet IP address outside that allowlist. Because the allowlist is enforced only for session-based login and not for API-key, OAuth, or header-based authentication, the request succeeds and the attacker operates with that user's privileges from an unauthorized location. No public exploit identified at time of analysis; the technique requires no special tooling beyond a standard HTTP client.