#vulns.co
/
mcp by GKData.io

← Back to Reports

Unsafe deserialization in a scientific-analysis project

The key takeaway is architectural: serialization formats that can revive executable behavior should not be an interchange boundary for untrusted artifacts.

Editorial decision card. This page links to a source-verified public disclosure and contains only original defensive analysis. It does not mirror upstream HTML, payloads, attachments, private submissions, or exploit steps.
Original severity
P1 source-reported; not rescored by vulns.co
Public source
bugcrowd crowdstream
Program / vendor
National Aeronautics and Space Administration (NASA) - Vulnerability Disclosure Program
Product / surface
GSFC HPLC Precision Analysis
Weakness
Insecure deserialization · CWE-502
Affected boundary
Cached analysis artifact processing
Disclosure date
2026-05-29
Public status checked
2026-09-10
Public attribution
kernely

What the evidence established

The public source attributes the condition to deserializing untrusted data and states that the risky format was replaced.

Why the impact was credible

The source reported code-execution risk in the analysis environment through unsafe object deserialization.

Durable engineering lesson

The key takeaway is architectural: serialization formats that can revive executable behavior should not be an interchange boundary for untrusted artifacts.

Control pattern

Avoid deserializing untrusted native objects; use data-only formats with schema validation and explicit integrity controls.

Primary public disclosure

Read the original source ↗

Upstream availability and wording can change. Public status was last checked 2026-09-10.

← Back to Reports