Executive brief
DeepSeek TUI is a terminal-based interface for interacting with AI models. A security flaw allows the application to be tricked into making unauthorized requests to internal network resources when processing specific IPv6 addresses. This could allow an attacker to access sensitive data or services on the local network that are normally restricted from outside access.
Technical details
A Server-Side Request Forgery (SSRF) vulnerability exists in the `fetch_url` tool of DeepSeek TUI. While the application attempts to validate hostnames against private IPv6 addresses, it fails to properly sanitize or block direct IPv6 literal inputs such as `http://[::1]`. An attacker can exploit this by providing a crafted URL to the tool, bypassing existing SSRF defenses to reach internal services on the loopback interface or local network. The vulnerability is rooted in `crates/tui/src/tools/fetch_url.rs` and has been addressed in version 0.8.26.
Affected products
- Hmbown deepseek-tui < 0.8.26
Timeline
- 2026-05-10: disclosed
- 2026-05-14: advisory: GitHub Advisory published
- 2026-05-28: other: NVD published
References
- https://api.github.com/users/JafarAkhondali
- https://github.com/JafarAkhondali
- https://api.github.com/users/JafarAkhondali/gists%7B/gist_id%7D
- https://api.github.com/users/JafarAkhondali/repos
- https://avatars.githubusercontent.com/u/11364402?v=4
- https://api.github.com/users/JafarAkhondali/events%7B/privacy%7D