What is CWE-913?
The product does not properly restrict reading from or writing to dynamically-managed code resources such as variables, objects, classes, attributes, functions, or executable instructions or statements.
Many languages offer powerful features that allow the programmer to dynamically create or modify existing code, or resources used by code such as variables and objects. While these features can offer significant flexibility and reduce development time, they can be extremely dangerous if attackers can directly influence these code resources in unexpected ways.
CWE-913 is a class-level weakness in MITRE’s Common Weakness Enumeration. Applicable platforms: Language: Not Language-Specific; Language: Interpreted.
Source: MITRE CWE (CWE-913 definition, reproduced verbatim). Counts and linkage below are Threadlinqs data.
Consequences
- Integrity — Execute Unauthorized Code or Commands
- Other, Integrity — Varies by Context, Alter Execution Logic
Source: MITRE CWE, common consequences.
How CWE-913 is exploited in the wild
Threadlinqs maps 4 CVEs to CWE-913, published between 2021-03-30 and 2026-10-08. 1 is listed in CISA’s Known Exploited Vulnerabilities catalog, the authoritative record of exploitation in the wild. By CVSS v3 severity the set splits into 2 critical, 1 high. The highest EPSS score in the set is 98.9% (CVE-2025-68613), the modelled probability of exploitation in the next 30 days. 16 tracked threats reference CWE-913 directly or through a CVE it covers; the most recent is “CVE-2026-103663: Ollama /api/pull path traversal (digestToPath) enables unauthenticated remote file write and root RCE” (2026-10-08). Affected products concentrate in N8n (2), Ollama (1), laverdet (1).
Vulnerabilities (CVEs)
All 4 CVEs mapped to CWE-913, CISA KEV first, then by CVSS score.
- CVE-2025-68613 — CISA KEV · CVSS 9.9 critical · EPSS 98.9% · published 2025-12-19
- CVE-2026-25049 — CVSS 9.9 critical · EPSS 1.5% · published 2026-02-04
- CVE-2021-21413 — CVSS 8 high · EPSS 0.7% · published 2021-03-30
- CVE-2026-103663 — EPSS 0.7% · published 2026-10-08
Affected vendors
Threat activity
16 tracked threats cite CWE-913:
- CVE-2026-103663: Ollama /api/pull path traversal (digestToPath) enables unauthenticated remote file write and root RCEHIGH
- Critical Type Confusion in isolated-vm ExternalCopy Enables Guest-to-Host Sandbox Escape and RCE (GHSA-864f-rcv7-6rh4)CRITICAL
- CVE-2026-34486: Apache Tomcat EncryptInterceptor Bypass Actively Exploited in SnowLight and AI-Assisted CampaignsHIGH
- AsyncAPI npm Supply Chain Attack: Pwn-Request GitHub Actions Compromise Deploys Miasma Tasking FrameworkCRITICAL
- Cordyceps: Systemic Cross-Workflow Privilege-Escalation Supply-Chain Flaw in GitHub Actions CI/CD Pipelines (Microsoft Azure Sentinel, Google ADK, Apache Doris, Cloudflare Workers SDK, PSF Black)CRITICAL
- AI Skill-Scanner Bypass — ClawHub, Cisco & Vercel Malicious-Skill Detectors Evaded via Truncation, .pyc Bytecode, Archive Indirection & Prompt Injection (Trail of Bits)HIGH
- forge-jsxy npm Supply Chain RAT — 22 Versions in 22 Days with Crypto Wallet Theft, WebRTC P2P Exfil & Cross-Platform Persistent Backdoor (OSV MAL-2026-3609, SafeDep)HIGH
- KnowledgeDeliver LMS ViewState Deserialization Zero-Day CVE-2026-5426 — Unauthenticated RCE via Shared ASP.NET machineKey + BLUEBEAM (Godzilla) Web Shell and Cobalt Strike BEACON Watering HoleCRITICAL
- durabletask PyPI Supply Chain Compromise (v1.4.1–1.4.3) — Microsoft-Published Azure Durable Functions SDK Trojanized w/ Cross-Cloud Credential Stealer + Linux Disk Wiper (TeamPCP / Mini Shai-Hulud)CRITICAL
- Agent Tesla LATAM Operation — 18-Month Credential Theft Campaign Against Chilean Enterprises via .NET Reactor 6.x Loader & aspnet_compiler.exe Process HollowingHIGH
- NGate Android NFC Relay Malware Variant - Trojanized HandyPay Banking App Campaign Targeting Brazil (2025-2026)HIGH
- Apache ActiveMQ OpenWire Deserialization RCE (CVE-2023-46604) — 6,400 Brokers Actively Exploited by HelloKitty, Kinsing, TellYouThePass and Andariel (Lazarus)CRITICAL
- CVE-2026-34197 — Apache ActiveMQ Jolokia Code Injection via Spring XML Context (CISA KEV)HIGH
- n8n Workflow Automation RCE via Expression Injection — CVE-2025-68613 (CVSS 9.9) Active Exploitation by Zerobot BotnetCRITICAL
- Seedworm (MuddyWater) Iranian MOIS APT Campaign Targeting U.S. Critical Infrastructure with Dindoor and Fakeset BackdoorsCRITICAL
- n8n Multi-CVE Vulnerability Cascade — Expression Sandbox Escape, Pyodide RCE, Arbitrary File Write, Command Injection (9 CVEs, 2× CVSS 10.0)CRITICAL
Mitigations
- Implementation / Input Validation: For any externally-influenced input, check the input against an allowlist of acceptable values.
- Implementation, Architecture and Design / Refactoring: Refactor the code so that it does not need to be dynamically managed.
Source: MITRE CWE, potential mitigations.
Detection methods (MITRE CWE)
- Fuzzing (effectiveness: High): Fuzz testing (fuzzing) is a powerful technique for generating large numbers of diverse inputs - either randomly or algorithmically - and dynamically invoking the code with those inputs. Even with random inputs, it is often capable of generating unexpected results such as crashes, memory corruption, or resource consumption. Fuzzing effectively produces repeatable test cases that clearly indicate bugs, which helps developers to diagnose the issues.
Source: MITRE CWE, detection methods. Threadlinqs detection rules for the threats above are Blue tier and higher.