CVE-2026-82447

Published: Ago 29, 2026 Last Modified: Ago 29, 2026
ExploitDB:
Other exploit source:
Google Dorks:
HIGH 8,7
Attack Vector: network
Attack Complexity: low
Privileges Required: low
User Interaction: none
Confidentiality: N/A
Integrity: N/A
Availability: N/A
HIGH 8,8
Attack Vector: network
Attack Complexity: low
Privileges Required: low
User Interaction: none
Scope: unchanged
Confidentiality: high
Integrity: high
Availability: high

Description

AI Translation Available

Skyvern before 1.0.45 contains a sandbox escape vulnerability in TextPromptBlock that renders prompts twice, first through a sandboxed Jinja environment and then through an unsandboxed environment. Attackers can inject malicious Jinja template syntax through workflow parameters or upstream block output to execute arbitrary code with server process privileges.

1336

Improper Neutralization of Special Elements Used in a Template Engine

Incomplete
Common Consequences
Security Scopes Affected:
Integrity
Potential Impacts:
Execute Unauthorized Code Or Commands
Applicable Platforms
Languages: Java, PHP, Python, JavaScript, Interpreted
Technologies: Not Technology-Specific, AI/ML, Client Server
View CWE Details
https://github.com/Skyvern-AI/skyvern
https://github.com/Skyvern-AI/skyvern/blob/v1.0.44/skyvern/forge/sdk/prompting.…
https://github.com/Skyvern-AI/skyvern/blob/v1.0.44/skyvern/forge/sdk/workflow/m…
https://github.com/Skyvern-AI/skyvern/commit/d723de621d5b3a340f3cc4d5b46bfe40a9…
https://www.vulncheck.com/advisories/skyvern-before-1.0.45-sandbox-escape-via-t…