CVE-2026-54769
A permanent, source-backed tracker page. Values are shown only when returned; explicit gaps and read failures remain different states.
What it is
ReportedLangroid is a framework for building large-language-model-powered applications. Versions prior to 0.65.2 are vulnerable to a critical Sandbox Escape leading to Remote Code Execution (RCE) in its `TableChatAgent` and `VectorStore` capabilities. When these agents evaluate LLM-generated tool messages with `full_eval=True`, they attempt to sandbox the execution by explicitly setting `locals` to an empty dictionary `{}` inside Python's `eval()` function. However, this relies on an incomplete understanding of Python's execution model.… Source description excerpt; complete tracked detail loads below.
NVD published: Jul 10, 2026 · Source record updated: Aug 8, 2026 · Tracker snapshot: Aug 8, 2026
What is affected
ReportedLangroid — Langroid. Product-level identification only; no affected-version conclusion is available from this field.
NVD published: Jul 10, 2026 · Source record updated: Aug 8, 2026 · Tracker snapshot: Aug 8, 2026
Urgency and basis
ReportedAttend · 339.6
Tracker decision tier from the evidence detailed below · Source record updated: Aug 8, 2026 · Tracker snapshot: Aug 8, 2026
Exploitation evidence
Not reportedNot reported by tracked sources.
What to do
ReportedA source remediation link is available. Validate that guidance against the affected product and your environment.
Source remediation ↗Tracked source remediation field · Source record updated: Aug 8, 2026 · Tracker snapshot: Aug 8, 2026
Evidence detail
Loading the full tracker evidence record…