CVE-2023-46248
Sourcegraph Cody 0.10.0 – 0.14.0
Raw vector
CVSS:3.1/AV:N/AC:L/PR:L/UI:R/S:C/C:H/I:H/A:HCVSS and EPSS are reproduced from their sources (NVD, FIRST EPSS). Risk Priority is our own derived reading, not an NVD score.
Summary
CVE-2023-46248 is a critical-severity External Control of System or Configuration Setting (CWE-15) vulnerability in Sourcegraph Cody. Its CVSS base score is 9.0 (Critical).
Operationally, exploitation aligns with the MITRE ATT&CK technique Winlogon Helper DLL (T1547.004); ranked in the top 35% of CVEs by exploit likelihood; it is not currently listed in the CISA KEV catalog; a public proof-of-concept is referenced.
This vulnerability is AI-related — categorised as Enterprise AI Assistants; in the Supply Chain and Deployment risk domain.
OWASP Top 10 for Web (2025)
EU & UK References
- 🇪🇺 ENISA EUVD: EUVD-2023-50478
Vulnerability Data
Cody is an artificial intelligence (AI) coding assistant. The Cody AI VSCode extension versions 0.10.0 through 0.14.0 are vulnerable to Remote Code Execution under certain conditions. An attacker in control of a malicious repository could modify the Cody configuration file…
more
`.vscode/cody.json` and overwrite Cody commands. If a user with the extension installed opens this malicious repository and runs a Cody command such as /explain or /doc, this could allow arbitrary code execution on the user's machine. The vulnerability is rated as critical severity, but with low exploitability. It requires the user to have a malicious repository loaded and execute the overwritten command in VS Code. The issue is exploitable regardless of the user blocking code execution on a repository through VS Code Workspace Trust. The issue was found during a regular 3rd party penetration test. The maintainers of Cody do not have evidence of open source repositories having malicious `.vscode/cody.json` files to exploit this vulnerability. The issue is fixed in version 0.14.1 of the Cody VSCode extension. In case users can't promptly upgrade, they should not open any untrusted repositories with the Cody extension loaded.
- CWE(s)
AI Security AnalysisAI
- AI Category
- Enterprise AI Assistants
- Risk Domain
- Supply Chain and Deployment
- OWASP Top 10 for LLMs 2025
- None mapped
- Classification Reason
- Cody is explicitly described as an artificial intelligence (AI) coding assistant, with the vulnerability affecting its VSCode extension, fitting the Enterprise AI Assistants category as it targets an AI-powered development tool.
Related Threats
MITRE ATT&CK Enterprise Techniques
CVEs Like This One
Affected Assets
Mitigating Controls
Control response
—
—
- 4 hardening rules · 2 OS baselines
—
Likely Mitigating Controls AI
Per-CVE control mapping for this CVE has not run yet; the list below is derived from the weakness types (CWEs) cited in the NVD entry.
The policy and procedures establish internal controls and change management for system configuration settings, reducing the feasibility of external unauthorized modifications.
Baseline configuration under change control directly prevents unauthorized external modification of system or configuration settings.
Requires approval, documentation, and security impact review of all configuration changes, directly preventing unauthorized external control of system settings.
Impact analysis of configuration changes reduces the risk of deploying settings that permit unauthorized external control.
Restricting changes to system and configuration settings prevents external entities from controlling those settings without approval.
Establishing, implementing, approving deviations from, and monitoring configuration settings directly prevents external or unauthorized control of system settings.
The plan defines processes for identifying and managing configuration items, preventing external unauthorized control of system settings.
Vulnerability scanners directly detect externally controllable or misconfigured settings using standardized checklists.
Mitigating Controls (NIST CSF 2.0) AI
Derived directly from the weakness types (CWEs) cited in the NVD entry via our AI-authored CWE→CSF cross-walk (authority under review) — links open the control.
Formal change and exception management catches externally driven configuration alterations.
Least-privilege access policies reduce unauthorized external modification of configuration values.
Hardened configuration baselines and change controls directly limit external manipulation of settings.
Mitigating Controls (ISO/IEC 27001:2022 Annex A) AI
Derived directly from the weakness types (CWEs) cited in the NVD entry via our AI-authored CWE→ISO cross-walk (authority under review) — links open the control.
A.5.15 only requires that access control policies and rules be defined and documented; it does not itself enforce, implement or configure those rules to stop external control of system/configuration settings, so the weakness remains fully exploitable.
A.5.18 governs the provisioning, review, and revocation of authorized access rights; it has no bearing on whether a system or configuration setting can be externally controlled by a user (CWE-15).
Documented operating procedures define how configuration changes must be performed.
Use of privileged utility programs is controlled to prevent misuse of configuration tools.
Privileged access rights reduce the number of users who can alter configuration.