THREATPORT/VULNERABILITY/CVE-2025-61260
CRITICAL SEVERITY

CVE-2025-61260

CVE-2025-61260 vulnerability

CVSS
9.8
EPSS
6.6%
EPSS pct
93th
Exploited
No known
A vulnerability was identified in OpenAI Codex CLI v0.23.0 and before that enables code execution through malicious MCP (Model Context Protocol) configuration files. The attack is triggered when a user runs the codex command inside a malicious or compromised repository. Codex automatically loads project-local .env and .codex/config.toml files without requiring user confirmation, allowing attackers to embed arbitrary commands that execute immediately.
Affected vendor
See references
Affected product
See references
CVSS vector
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
Weakness type (CWE)
Published
2026-04-14

Risk analysis

Based on its CVSS vector, this vulnerability is exploitable over the network, low attack complexity, requiring no privileges, no user interaction. Successful exploitation leads to high impact to confidentiality, high impact to integrity, high impact to availability.

Its EPSS score of 6.6% reflects a lower probability of exploitation activity in the wild over the next 30 days, placing it above 93% of all scored CVEs.

Check your external exposure to CVE-2025-61260

  • Confirm whether the affected software is running on any internet-facing host or subdomain.
  • Match discovered service fingerprints against the version affected by CVE-2025-61260.
  • Prioritise remediation if the asset is public, business-critical, or near authentication/payment flows.

How ThreatPort helps with CVE-2025-61260

ThreatPort's external attack surface scanner checks your internet-facing assets — agentlessly, from the outside in — for the exposed services and versions that CVE-2025-61260 affects, then prioritises the finding by real risk using CISA KEV and FIRST.org EPSS.

Run a free external exposure scan

Are you exposed to CVE-2025-61260?

Run an instant, non-intrusive external scan to check your attack surface for CVE-2025-61260.

Start free scan

No credit card. Agentless.

Related vulnerabilities

CRITICAL
CVE-2018-16283
Wechat Brodcast vulnerability
CRITICAL
CVE-2018-17552
Navigate Cms vulnerability
CRITICAL
CVE-2018-15379
Prime Infrastructure vulnerability