# CVE-2026-51869: DB-GPT sandbox API fallback to LocalRuntime code execution

Source: https://junglewise.ai/threats/cve-2026-51869-db-gpt-v0-8-0-sandbox-api-silently-falls-back-to-localruntime-and  
Publisher: Junglewise Threat Intelligence, https://junglewise.ai/threats  
Published: 2026-09-30. Updated: 2026-09-30  
License: CC BY 4.0 with a link back

CVE-2026-51869 · Severity: info · Published 2026-09-30

DB-GPT is an AI agent platform that executes code based on natural language requests. Version 0.8.0's sandbox API contains a flaw where it silently falls back from containerized execution (Docker, Podman, Nerdctl) to running code directly on the host system. An attacker can trigger arbitrary code execution on the host through the web UI by crafting prompts that the AI agent converts into code.

Technologies: [Eosphoros-Ai DB-GPT](https://junglewise.ai/threats/technologies/db-gpt). Vendors: [Eosphoros-Ai](https://junglewise.ai/threats/vendors/eosphoros-ai).

## Quick answers

**Is CVE-2026-51869 exploited in the wild?** No. Junglewise Threat Intelligence has no report of CVE-2026-51869 being exploited in the wild.

**How severe is CVE-2026-51869?** Informational severity, published 30 September 2026.

**Which products and versions are affected?** Eosphoros-Ai DB-GPT 0.8.0.

**Is there a fix for CVE-2026-51869?** No fixed version is recorded yet. Check the vendor's advisory in the references below. If you run it, Junglewise drafts a mitigation that fits your setup and applies it once you approve, with senior security engineers behind it.

**How is CVE-2026-51869 attacked?** It is attacked over the network.

**Am I affected by CVE-2026-51869?** Check whether you run Eosphoros-Ai DB-GPT at an affected version. Junglewise does that check for you around the clock: connect your code, cloud and domains, and it matches new vulnerabilities like this one against what you actually run and tells you when one reaches you.

## Technical details

The vulnerability is an incorrect access control flaw in runtime_factory.py:RuntimeFactory.create() where a fallback chain attempts containerized runtimes but lacks proper failure handling, ultimately executing code via LocalRuntime on the host when containerization is unavailable or misconfigured. This allows unauthenticated or semi-trusted users to achieve remote code execution through the web UI without the expected sandbox isolation. The issue affects DB-GPT v0.8.0 and stems from missing validation that sandboxing was actually achieved.

## Affected products

- eosphoros-ai DB-GPT 0.8.0

## Timeline

- 2026-05-22: disclosed: Issue reported on GitHub
- 2026-09-30: advisory: CVE-2026-51869 published

## References

- [https://gist.github.com/Ro1ME/884fbbd589998b81fa05c20ac205569f](https://gist.github.com/Ro1ME/884fbbd589998b81fa05c20ac205569f)
- [https://github.com/eosphoros-ai/DB-GPT/issues/3082](https://github.com/eosphoros-ai/DB-GPT/issues/3082)

## Related threats

- [eosphoros-ai DB-GPT remote code execution in HuggingFace adapter](https://junglewise.ai/threats/cve-2026-123019-a-weakness-has-been-identified-in-eosphoros-ai-db-gpt-up-to-0-8)
- [eosphoros-ai DB-GPT code injection in CodeAction](https://junglewise.ai/threats/cve-2026-123018-a-security-flaw-has-been-discovered-in-eosphoros-ai-db-gpt-up-to)
- [DB-GPT directory traversal in use_example_file](https://junglewise.ai/threats/cve-2026-51863-db-gpt-v0-7-5-and-0-8-0-contains-directory-traversal-in-use)
- [DB-GPT hardcoded credentials in default Docker configuration](https://junglewise.ai/threats/cve-2026-77933-db-gpt-through-0-8-2-contains-a-hardcoded-credentials)
- [DB-GPT unauthenticated remote code execution](https://junglewise.ai/threats/cve-2026-77932-db-gpt-through-0-8-2-contains-an-unauthenticated-remote-code)