Loading vulnerability…
CVE-2026-8838 is a code injection vulnerability in amazon-redshift-python-driver, distributed as the PyPI package redshift-connector, affecting versions before 2.1.14. The vector_in() function uses Python's eval() to deserialize untrusted vector data received from a server over the PostgreSQL wire protocol. A rogue server or man-in-the-middle attacker can return crafted query results that execute arbitrary Python code in the client process.
Mallory correlates every CVE against your assets, your vendors, and active adversary campaigns. Know which vulnerabilities matter for you, not just which ones are loud.
What it means. What to do now. Patch path, mitigations, and the assume-compromise checklist.
What an attacker gets, and what they’ve been doing with it.
If you can’t patch tonight, do this now.
Patch, then assume compromise.
1 valid exploit after Mallory filtered fakes, detection scripts, and README-only repos (1 hidden).
This repository is a small standalone proof-of-concept for CVE-2026-8838, described as a remote code execution issue in the Amazon Redshift Python driver caused by unsafe use of eval() on server-supplied data in vector deserialization. The repo contains two Python scripts and two Markdown documents: poc_server.py implements a fake Redshift-like TCP server that listens on 127.0.0.1:5439 and returns a crafted response containing Python code; poc_client.py simulates the vulnerable client behavior by connecting to that server, receiving the forged response, and passing it to eval(), then contrasts that behavior with a safer ast.literal_eval()-based implementation. The included demonstration payload is __import__('os').popen('whoami').read(), which shows client-side command execution and output capture. The exploit capability is therefore arbitrary code execution on the machine running the vulnerable client, assuming the attacker can operate a rogue server or intercept/modify Redshift traffic as a MitM. README.md documents the vulnerability, impact, affected versions, and usage steps, while MITIGATION.md provides defensive guidance such as upgrading to version 2.1.14+, enforcing TLS certificate validation, and restricting network access. Overall, this is an operational educational PoC rather than a full weaponized exploit framework.
Products and vendors Mallory has correlated with this vulnerability. Open in Mallory to drill down to specific CPE configurations and version ranges.
Vendor-confirmed product mapping. Mallory continuously reconciles this list against your asset inventory.
12 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.
An unspecified network-accessible vulnerability requiring no privileges or user interaction, with high confidentiality, integrity, and availability impacts according to the supplied CVSS vectors. Exploits are available, and a patch publication date of July 1, 2026 is listed. The affected product and technical cause are not identified.
A remote code execution vulnerability in amazon-redshift-python-driver before version 2.1.14 caused by unsafe use of Python eval() on server-received data in vector_in(), allowing a rogue server or man-in-the-middle attacker to execute arbitrary code on the client.
A critical remote code execution vulnerability in the official Amazon Redshift Python driver caused by use of eval() on untrusted server-supplied vector data, allowing a rogue server or man-in-the-middle to execute arbitrary Python code on the client.
Query your assets running an affected version, and investigate the blast radius.
Every observed campaign linking this CVE to a named adversary.
Malware families riding this exploit, with evidence and IOCs.
YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.
Cross-references every affected SKU, including bundled OEM variants.
Community discussion across Reddit, Mastodon, and other social sources.