Vulnerability · Published Aug 29, 2026 · Assigned by VulnCheck
CVE-2026-82447Skyvern before 1.0.45 Sandbox Escape via TextPromptBlock
High 8.8
In plain English
A high-severity security flaw in skyvern from Skyvern-AI. There are no confirmed reports of exploitation so far. The chance it is exploited in the next 30 days is estimated at 0.8% (higher than 56% of all known vulnerabilities).
Severity (CVSS 3.1)8.8High
Exploit odds (30 days)0.8%top 44.0% of all CVEs
Exploited in the wildNo reportsCISA SSVC: none
Attack toolingNoneno public tooling known
Official description
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.
Affected products (as reported)
| Vendor | Product | Versions |
|---|---|---|
| Skyvern-AI | skyvern | 0.2.1 – before 1.0.45 |
Weakness: CWE-1336 · Vector: CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
References
- Patch Commit patch
- https://github.com/Skyvern-AI/skyvern product
- Vulnerable code: block.py (sandboxed first render, non-sandboxed second render) technical-description
- Vulnerable code: prompting.py (PromptEngine uses a plain jinja2.Environment) technical-description
- VulnCheck Advisory: Skyvern before 1.0.45 Sandbox Escape via TextPromptBlock third-party-advisory