MISP Guard Case-Sensitive Policy Bypass Allows Unauthorized Feed Data Forwarding

MISP Guard, a proxy-based security filter for MISP feed distribution, contains a case-sensitivity flaw in multiple policy enforcement checks. The functions check_required_taxonomy_exists, check_required_taxonomy_tags_allowed, check_blocked_attribute_categories, and check_blocked_analyst_data_sharing_groups_uuids performed case-sensitive string comparisons when matching event tags, attribute categories, and sharing-group UUIDs against configured allow/block lists. An attacker with the ability to create or modify events in a MISP instance (or a connected instance that pushes events through the guard) could craft events using different letter case in taxonomy tag names, attribute categories, or sharing-group UUIDs to bypass the configured security policies. For example, if the policy requires the 'tlp' taxonomy with only 'tlp:clear' allowed, an event tagged 'TLP:AMBER' would not be recognized as belonging to the 'tlp' taxonomy and would therefore not be checked against the allowed-tags list, resulting in the event being forwarded to feed consumers. Similarly, a blocked attribute category such as 'Person' would not match a configured block entry of 'person', allowing the attribute to pass through the filter. The security impact is unauthorized disclosure of threat-intelligence data that the operator intended to restrict. Events carrying non-allowed TLP (Traffic Light Protocol) markings, blocked attribute categories, or blocked sharing-group associations are forwarded to downstream feed consumers in violation of the configured distribution policy. This undermines the confidentiality controls that MISP Guard is designed to enforce. The vulnerability is trivially exploitable: the attacker only needs to alter the letter case of a tag, category, or UUID string. No authentication beyond normal MISP event-creation privileges is required, and no user interaction is needed.