Unauthenticated Arbitrary File Write in Feast Feature Server

A vulnerability has been identified in the Feast Feature Server’s `/save-document` endpoint that allows an unauthenticated remote attacker to write arbitrary JSON files to the server's filesystem. Although the system attempts to restrict file locations, these protections can be bypassed, enabling an attacker to overwrite vital application configurations or startup scripts. Because this flaw requires no credentials or special privileges, any attacker with network access to the server can potentially compromise the integrity of the system. This could lead to unauthorized system modifications, denial of service through disk exhaustion, or potential remote code execution.

Casky was already ahead

This CVE exploits attack patterns that Casky's 283matched skills already investigate — long before this vulnerability was disclosed. Claude's reasoning model maps these techniques to MITRE ATT&CK, so practitioners who ran these skills have already seen the threat behaviour in their findings.

Analysis

CVE-2026-23537 represents a critical authorization bypass vulnerability in Feast Feature Server's `/save-document` endpoint that permits unauthenticated attackers to write arbitrary JSON files to the server filesystem. The vulnerability's 9.1 CVSS score reflects its severity: any networked attacker can exploit this without credentials to overwrite application configurations, startup scripts, or other sensitive files, potentially leading to complete system compromise. Organizations deploying Feast—a feature management platform commonly used in ML/AI pipelines—face immediate risk of data exfiltration, lateral movement, or denial of service if this endpoint is internet-exposed.

Casky's 283 matched skills leverage Claude's extended reasoning to identify the attack chain spanning MITRE techniques TA0004 (Privilege Escalation) and TA0006 (Credential Access). Practitioners using Casky would detect multiple behavioral indicators: CWE-862 (Missing Authorization) patterns in HTTP requests to `/save-document` lacking authentication tokens, followed by file writes to unexpected directories outside intended sandboxes. Claude's reasoning engine correlates these signals—unauthenticated POST requests with JSON payloads, filesystem write operations to config directories, and subsequent process execution from modified startup scripts—into a coherent attack narrative. The platform would flag suspicious file modification timestamps clustering around unauthorized API calls and alert on path traversal attempts (../ sequences) in document parameters, enabling defenders to identify exploitation before attackers establish persistence.

Live Threat IntelligenceComing soon

Casky Risk Score

—

EPSS Probability

—

Shodan Exposure

—

GreyNoise Signal

—

Composite risk scoring from EPSS, CISA KEV, Shodan, and GreyNoise — 21 security APIs correlated into a single Casky Risk Score. Coming in Casky Pro. Join early access →

283 Casky skills cover this attack pattern

These skills use Claude AI's reasoning model to surface findings in the same attack categories as CVE-2026-23537.

abusing-dpapi-for-credential-access

red teaming · high

Run

abusing-shadow-credentials-for-privesc

red teaming · high

Run

analyzing-cloud-storage-access-patterns

cloud security · low

Run

MITRE ATT&CK Techniques

TA0004 TA0006

Investigate the attack patterns behind CVE-2026-23537

Casky has 283 skills that investigate the attack patterns behind CVE-2026-23537. Run one and get CVSS-scored findings in 3 minutes.

Run the skill that detects this →

X Instagram LinkedIn