Ginlix-ai LangAlpha
LangAlpha through 2026.09.07 Indirect Prompt Injection RCE via Bash and ExecuteCode
LangAlpha through 2026.09.07 contains an indirect prompt injection vulnerability that allows remote attackers to execute shell commands or Python by planting instructions in web content retrieved by the agent. Because Bash and ExecuteCode lack an allow-list and human-in-the-loop approval, attackers can run code in the workspace sandbox and exfiltrate vault secrets from _internal/.vault_secrets.json.
First public PoC
None yetno public PoC found
Public PoCs00 on GitHub
CVSS8.8high · scored by CNA · CWE-1427
CISA KEV
Not listednot known exploited
Affected versions and mitigation
From the CVE record and the vendor's references.Affected
ginlix-ai LangAlpha
- <= 2026.09.07
Mitigation
The record names no fix. Check the vendor advisory below.
Timeline
Days relative to CVE publicationNo public PoC found yet; not in CISA KEV.
PoC repositories
Found on GitHub.No PoC repositories found yet.
Other public exploits
Exploit-DB, Metasploit and nuclei count as public PoCs. Exploit-tagged references from the CVE record and NVD are links only. None of these are analysed by us.| Source | Title | Published | Found by us |
|---|---|---|---|
| CVE record (tagged exploit) | Lasso.security Advisory | unknown | 14 Nov 2026 |