Frappe Framework CVE-2026-41581

No special conditions are required for exploitation - the CVSS 4.0 vector (AV:N/AC:L/AT:N/PR:N/UI:N) confirms this is a remote, unauthenticated attack against a network-accessible endpoint with no special attack prerequisites. The blog module must be deployed and the get_blog_list function must be reachable via HTTP/HTTPS, which is standard for Frappe deployments using the blog feature. What limits exploitation: the impact is bounded to Low confidentiality and Low integrity (no full database dump, no RCE, no availability impact per CVSS VC:L/VI:L); the blog module must be in use (deployments that have disabled or never configured the blog feature are not exposed through this specific vector); and patched versions 15.106.0 and 16.16.0 fully remediate the flaw.

Real-world risk is moderate-to-low for most operators. The CVSS 4.0 base score of 6.9 reflects a meaningful but bounded impact: confidentiality and integrity are both rated Low (VC:L/VI:L), with no availability impact and no scope change to subsequent systems (SC:N/SI:N/SA:N). EPSS at 0.02% in the 5th percentile confirms that automated or mass exploitation has not been observed and is statistically unlikely in the near term. This is not listed in CISA KEV, and no public proof-of-concept code has been identified. The primary concern is that the endpoint is unauthenticated and network-accessible (AV:N/PR:N), meaning any internet-facing Frappe deployment is in scope without any preconditions beyond connectivity. SQL injection impact is bounded here - partial data leakage and limited data manipulation - rather than full database dump or remote code execution, which limits its severity relative to typical high-CVSS SQLi.

An unauthenticated attacker sends a crafted HTTP GET or POST request to the Frappe blog listing endpoint, injecting SQL syntax into a query parameter handled by get_blog_list. The injected payload manipulates the underlying SQL query, allowing the attacker to enumerate database schema, extract partial application data, or perform limited data writes - all without requiring any account or session on the target application. No public proof-of-concept exploit code has been identified at time of analysis.