LLM01
Prompt Injection
133 CVEsCRITICALCVE-2026-54769CVSS 10.0Langroid: prompt injection to RCE via broken eval() sandbox
CRITICALCVE-2026-61447CVSS 10.0PraisonAI: RCE via unsandboxed LLM code execution
+ 130 more CVEs mapped to this control
LLM02
Sensitive Information Disclosure
346 CVEsCRITICALCVE-2026-61447CVSS 10.0PraisonAI: RCE via unsandboxed LLM code execution
CRITICALCVE-2026-34938CVSS 10.0praisonaiagents: sandbox bypass enables full host RCE
CRITICALCVE-2026-33663CVSS 10.0n8n: member role steals plaintext HTTP credentials
+ 343 more CVEs mapped to this control
LLM03
Supply Chain Vulnerabilities
560 CVEsCRITICALCVE-2026-111057CVSS 10.0Mercury: regex blocklist bypass allows unauthenticated RCE
CRITICALCVE-2026-94603CVSS 10.0Podman: checkpoint images override sandbox on run
+ 557 more CVEs mapped to this control
LLM04
Data and Model Poisoning
200 CVEsCRITICALCVE-2025-53002CVSS 9.8LLaMA-Factory: RCE via unsafe checkpoint deserialization
CRITICALCVE-2026-55450CVSS 9.3Langflow: unauthenticated upload → DoS + path disclosure
+ 197 more CVEs mapped to this control
LLM05
Improper Output Handling
316 CVEsCRITICALCVE-2026-35307CVSS 10.0Oracle Coherence: unauthenticated RCE, CVSS 10.0
CRITICALCVE-2025-15379CVSS 10.0MLflow: RCE via unsanitized model dependency specs
+ 313 more CVEs mapped to this control
LLM06
Excessive Agency
836 CVEsCRITICALCVE-2026-33663CVSS 10.0n8n: member role steals plaintext HTTP credentials
CRITICALCVE-2026-46695CVSS 10.0Boxlite: read-only bypass enables host code execution
+ 833 more CVEs mapped to this control
LLM07
System Prompt Leakage
388 CVEsCRITICALCVE-2026-34938CVSS 10.0praisonaiagents: sandbox bypass enables full host RCE
CRITICALCVE-2025-53767CVSS 10.0Azure OpenAI: SSRF EoP, no auth required (CVSS 10)
CRITICALCVE-2025-71338CVSS 10.0Flowise: unauthenticated file write enables RCE
+ 385 more CVEs mapped to this control
LLM08
Vector and Embedding Weaknesses
366 CVEsCRITICALCVE-2025-14931CVSS 10.0smolagents: RCE via pickle deserialization in executor
CRITICALCVE-2026-34938CVSS 10.0praisonaiagents: sandbox bypass enables full host RCE
CRITICALCVE-2026-54769CVSS 10.0Langroid: prompt injection to RCE via broken eval() sandbox
+ 363 more CVEs mapped to this control
LLM09
Misinformation
20 CVEsCRITICALCVE-2025-61260CVSS 9.8OpenAI Codex CLI: RCE via malicious MCP config files
CRITICALCVE-2024-39236CVSS 9.8Gradio: code injection via component metadata (CVSS 9.8)
+ 17 more CVEs mapped to this control
LLM10
Unbounded Consumption
126 CVEsCRITICALCVE-2026-57116CVSS 9.8PraisonAI: unauthenticated AgentOS chat API drives agents
CRITICALCVE-2026-57118CVSS 9.8PraisonAI agents: unauthenticated access to agent endpoints
HIGHCVE-2026-92954CVSS 8.6vm2: unhandled host Promise rejection crashes Node host
+ 123 more CVEs mapped to this control
Download Full Evidence Pack
Get the complete OWASP LLM Top 10 evidence pack with all CVE-to-control mappings, rationale, and audit-ready documentation. Exportable as CSV.
Get Evidence Pack