Veeam, Terraform MCP, Django Patch 11 Critical Flaws Including CVSS 10.0
Three major software vendors — Veeam, HashiCorp, and the Django Software Foundation — released patches on August 5, 2026, addressing 11 vulnerabilities across Terraform MCP Server, Veeam Service Provider Console (VSPC), and Django. Three flaws stand out as critical: an unauthenticated credential-exposure bug in Veeam's service provider console (CVSS 9.5), a maximum-scored cross-tenant token-reuse vulnerability in HashiCorp's Terraform MCP Server (CVSS 10.0), and a file-write flaw in GeoDjango's spatial lookups that can lead to remote code execution on certain configurations. None of the CVEs appear in CISA's Known Exploited Vulnerabilities catalog, and no public proof-of-concept has surfaced as of publication.
Veeam Service Provider Console, used by hosting providers and MSPs to manage multi-tenant backups, received four fixes in build 9.3.0.35057, released July 29 and detailed in an August 4 security bulletin. The most severe is CVE-2026-58073 (CVSS 9.5), which lets an unauthenticated attacker impersonate a managed agent and harvest its credentials — a high-complexity but no-login path that makes exposed credentials a priority concern. Operators should verify any potentially compromised agent identities against known exposures using a email breach checker and rotate affected credentials immediately. The second critical bug, CVE-2026-58072 (CVSS 9.0), is an arbitrary file write enabling remote code execution with low-privilege access. Two additional high-severity flaws — CVE-2026-58067 (unauthenticated memory-exhaustion DoS) and CVE-2026-58071 (a brief session-window privilege escalation to Portal Administrator) — affect VSPC 9.2.1.33875 and all earlier version 9 builds. This marks the second critical VSPC patch cycle in roughly three months, following May's CVE-2026-32998 (CVSS 9.4) remote code execution fix.
HashiCorp's Terraform MCP Server, which connects AI assistants to Terraform via the Model Context Protocol, carries the most alarming score: a CVE rated CVSS 10.0 for a cross-tenant flaw that allows one user's Terraform token to be reused for subsequent users' requests. The bug affects Streamable HTTP transport rather than stdio, and operators running Terraform MCP Server should upgrade immediately to version 1.1.0. Because AI assistant integrations are increasingly exposed to network-borne attacks, administrators should also confirm their MCP endpoints are properly encrypted by running an SSL/TLS checker against the relevant services.
The Django patch covers a flaw in GeoDjango's spatial lookups that can write a file to disk and, on certain configurations, execute code. The documented attack path requires a staff user with view permission on a registered model containing a spatial field — so it is not exploitable from the public internet, but it remains a serious concern for multi-author Django deployments. Affected versions are patched by upgrading to Django 6.0.8 or 5.2.17. For environments already potentially exposed, a broader privacy checkup can help identify lingering configuration weaknesses across the deployment.