2 vulnerabilities across 2 products scored HIGH or above on August 29, 2026.

  • 🟠 HIGH: 2

Exploit Status Upgrades

The following CVEs from previous bulletins have been upgraded based on new exploit intelligence:

  • [UPGRADED] CVE-2026-65641 (veeam/one) — F1: exploitable → functional, AAS: 10.4 → 12.4 (HIGH → CRITICAL). Originally in 2026-08-26 bulletin.

🟠 [HIGH] omnivore-app/omnivore

1 CVE | CVSS 4.0: 9.3 | AAS 10.2

  • cpe:2.3:a:omnivore-app:omnivore:*:*:*:*:*:*:*:* (< abf53d6)

Omnivore, the open-source read-it-later platform by omnivore-app, is affected by one high-severity vulnerability. CVE-2026-82454 (CVSS 9.3) is a critical authentication bypass in the Omnivore API’s Apple sign-in flow, where the decodeAppleToken function trusts the algorithm specified in an attacker-supplied JWT header. An attacker can forge authentication tokens by switching the algorithm to HS256 and signing with Apple’s publicly available RSA public key, gaining unauthorized access to any account that uses Apple sign-in. This is a classic algorithm confusion attack enabled by the use of jsonwebtoken v8, which does not enforce key-algorithm compatibility.

Organizations self-hosting Omnivore instances with Apple sign-in enabled should treat this as an urgent priority. Update immediately to a build that includes commit abf53d6 or later. If patching is not immediately possible, disable Apple sign-in as a temporary mitigation and audit authentication logs for anomalous token activity. Refer to the vendor repository at https://github.com/omnivore-app/omnivore for patch details.

Vendor Advisory

🟠 [HIGH] skyvern-ai/skyvern

1 CVE | CVSS 4.0: 8.7 | AAS 9.1

  • cpe:2.3:a:skyvern-ai:skyvern:*:*:*:*:*:*:*:* (< 1.0.45)

Skyvern, the AI-powered browser automation framework by skyvern-ai, is affected by one high-severity vulnerability. CVE-2026-82447 (CVSS 8.7) is a server-side template injection flaw in TextPromptBlock where prompts are rendered twice, first through a sandboxed Jinja environment and then through an unsandboxed one. An attacker who can control workflow parameters or upstream block output can inject malicious Jinja template syntax that survives the first pass and executes as arbitrary code with full server process privileges during the second, unsandboxed render.

Teams running Skyvern for workflow automation should upgrade to version 1.0.45 or later immediately, as this vulnerability is exploitable and provides a direct path to remote code execution. If immediate patching is not feasible, restrict access to workflow creation and audit existing workflows for untrusted parameter inputs. Refer to the vendor repository at https://github.com/Skyvern-AI/skyvern for patch details.

Vendor Advisory