The vulnerability stems from the handleFilteringSetRules function accepting text/plain content without proper CSRF protections. The pre-patch implementation directly read plain text rules from the request body (io.ReadAll) and split them, making it vulnerable to CSRF attacks via malicious links. The fix introduced JSON validation (filteringRulesReq struct) and Content-Type checks (via WriteTextPlainDeprecated), indicating the original text/plain processing was the vulnerability source. The function's direct connection to custom filtering rules modification and the security patch's focus on this endpoint confirm its role in the vulnerability.