// Security Advisory Triage
Langflow RCE via decorator injection, fixed in 1.10.1
GitHub advisories — pip · published 2026-10-06 18:27 UTC
WHAT CHANGEDGitHub published CVE-2026-51886 on 5 October 2026, rating it high severity. Langflow's validate_code function allows authenticated remote code execution through malicious decorator injection. Version 1.10.1 patches the vulnerability.
WHY IT MATTERSAn authenticated attacker can execute arbitrary code on the server by injecting malicious decorators through validate_code. This is remote code execution with authentication as the only barrier.
WHAT TO DOUpgrade any Langflow installation to 1.10.1 or later. If you cannot upgrade immediately, audit who has authentication credentials and restrict access to the validate_code endpoint until patched.
Model-reported confidence 95% (not measured accuracy)
Primary source ↗
written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
+ghsa_id: GHSA-w584-2h2r-2hvf +summary: Langflow: Title Authenticated Remote Code Execution in validate_code via Malicious Decorators Description +severity: high +published_at: 2026-10-05T22:31:00Z +cve_id: CVE-2026-51886 +vulnerabilities[].package.name: langflow +vulnerabilities[].first_patched_version: 1.10.1
HOW IT WAS JUDGED
Triage signals below are model-reported assertions, preserved verbatim; they are not independent arithmetic verification.
| triage verdict | MATERIAL at 0.92, needed 0.5 |
|---|---|
| triage reasoning | Langflow is an AI agent framework in dependency trees. RCE via decorators is exploitable by authenticated users with framework access. |
| signals | AI framework package (Langflow) · authenticated remote code execution · known fixed version published · high severity rating · validate_code function targeted · malicious decorator injection vector |
| suppression | — |
| triage model | anthropic/claude-haiku-4.5 ($0.00000) |
| writer model | anthropic/claude-sonnet-4.5 ($0.00000) |