CVE-2026-4965
Raw vector
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:L/VI:L/VA:L/SC:N/SI:N/SA:N/E:P/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:XSummary
CVE-2026-4965 is a medium-severity Code Injection (CWE-94) vulnerability. Its CVSS base score is 6.9 (Medium).
Operationally, exploitation aligns with the MITRE ATT&CK technique JavaScript (T1059.007); ranked at the 46th percentile by exploit likelihood (below the median); it is not currently listed in the CISA KEV catalog.
This vulnerability is AI-related — categorised as AI Agent Protocols and Integrations; in the LLM/Generative AI Risks risk domain.
The strongest mitigations our analysis identified map to SA-11 (Developer Testing and Evaluation) and SI-10 (Information Input Validation) — see the control section below for these in your framework.
Deeper analysis AI-assisted summary
Synthesised by an AI model from the NVD description and linked references — a reading aid, not an authoritative source.
CVE-2026-4965 is a code injection vulnerability stemming from an incomplete fix for CVE-2025-6101 in the letta-ai letta framework version 0.16.4. It affects the resolve_type function within the file letta/functions/ast_parsers.py, where improper neutralization of directives in dynamically evaluated code (classified under CWE-94 and CWE-95) allows malicious input to be processed unsafely. The issue carries a CVSS v3.1 base score of 7.3 (AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:L), indicating high severity due to its network accessibility and lack of prerequisites.
Remote attackers can exploit this vulnerability without authentication or user interaction by manipulating inputs to the affected function, potentially leading to arbitrary code execution or directive injection during dynamic code evaluation. Successful exploitation grants low-level impacts on confidentiality, integrity, and availability, enabling data leakage, modification of application state, or denial of service within the letta framework's parsing context.
Advisories from VulDB (ctiid.353842, id.353842, submit.777654) document the issue, with a public exploit available via a GitHub Gist (https://gist.github.com/YLChen-007/fc09bc447a73bba526c1642d9ce73ca5). No patches or vendor responses are reported, as the letta-ai team was contacted early but did not reply; security practitioners should avoid version 0.16.4 and monitor for updates.
The exploit's public availability increases the risk of active exploitation, particularly in environments using letta-ai for AI-related dynamic code processing.
OWASP Top 10 for Web (2025)
EU & UK References
- 🇪🇺 ENISA EUVD: EUVD-2026-16736
Vulnerability Data
A vulnerability was detected in letta-ai letta 0.16.4. This issue affects the function resolve_type of the file letta/functions/ast_parsers.py of the component Incomplete Fix CVE-2025-6101. Performing a manipulation results in improper neutralization of directives in dynamically evaluated code. The attack can…
more
be initiated remotely. The exploit is now public and may be used. The vendor was contacted early about this disclosure but did not respond in any way.
- CWE(s)
AI Security AnalysisAI
- AI Category
- AI Agent Protocols and Integrations
- Risk Domain
- LLM/Generative AI Risks
- OWASP Top 10 for LLMs 2025
- None mapped
- Classification Reason
- Matched keywords: ai
Related Threats
MITRE ATT&CK Enterprise Techniques
CVEs Like This One
Affected Assets
Mitigating Controls
Control response
—
—
—
V1.3.1V1.3.2
Mitigating Controls (NIST 800-53 r5) AI
Developer testing and evaluation finds code paths that accept and execute externally influenced strings.
Input validation directly stops untrusted data from being used to construct executable code without neutralization.
Least privilege limits the damage an injected code fragment can perform once executed.
Requiring documented secure development standards and tools enforces use of safe code-generation APIs and escaping.
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.
PR.PS-06's SDLC practices directly target injection flaws via secure coding and testing (mostly), yet as a single broad outcome it leaves many code-generation specifics unaddressed (partial).
PR.DS-10 protects runtime data confidentiality/integrity but has no bearing on neutralizing externally influenced input during code generation, so neither direction shows any preventive effect.
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.
Security testing in development can detect eval injection vulnerabilities before deployment.
Secure development life cycle mandates input validation and safe coding practices that directly prevent eval injection.
Application security requirements include rules against dynamic code execution of untrusted input.
Secure architecture principles discourage unsafe dynamic evaluation constructs.
Banning unapproved code samples and unauthenticated web services, combined with secure-coding standards and SAST, prevents the dynamic generation or inclusion of attacker-supplied code.
Separation of environments limits the blast radius if eval injection occurs in non-production.