Advisories · Weakness
CWE-1427.
Every Advisory and CVE whose sources give the weakness CWE-1427, newest first.
- CVSS 8.7 High
OpenHuman: Indirect prompt injection via background email sync executes Composio actions without user approval
- tinyhumansai openhuman
- CWE-1427
- CVSS 8.3 High
OpenHuman: MCP tool description injection: untrusted server descriptions execute in agent LLM context without sanitization
- tinyhumansai openhuman
- CWE-1427
- CVSS 5.9 Medium
Secret masking bypass in the environment tool in Strands Agents Tools
- amazon strands-agents-tools
- CWE-1427
- CVSS 8.7 High
LangAlpha through 2026.09.07 Indirect Prompt Injection RCE via Bash and ExecuteCode
- ginlix-ai LangAlpha
- CWE-1427
- CVSS 8.7 High
MoviePilot 3.1.5 Command Injection via ExecuteCommandTool
- jxxghp MoviePilot
- Command injection
- CWE-1427
- CVSS 7.3 High
Improper Neutralization of Input Used for LLM Prompting in GitLab
- GitLab
- CWE-1427
- CVSS 8.7 High
Vibe-Trading LLM-callable tools permit command execution, code injection, and SSRF
- Command injection
- Code injection
>=0.1.0, <0.1.7 · fixed in 0.1.7
GHSA-jqmf-mx4f-hfr6 · CVE-2026-114107 - CVSS 5.4 Medium
Microsoft Edge for iOS Spoofing Vulnerability
- Microsoft Edge (Chromium-based)
- Microsoft Edge for iOS
- CWE-1427
- CVSS 9.2 CriticalCritical · CVSS 9+
Consent bypass in python_repl tool via batch kwargs forwarding in Amazon Strands Agents Tools
- Amazon strands-agents-tools
- CWE-1427
- CVSS 6.4 Medium
Context7 2.1.2 Prompt Injection via Custom AI Instructions
- Uptash Context7
- CWE-1427
- CVSS 4.3 Medium
HCL AION is affected by multiple security vulnerabilities.
- HCL Software AION
- CWE-1427
- CVSS 7.5 High
Prompt injection bypasses shell tool consent gate in Strands Agents Tools
- AWS strands-agents-tools
- CWE-1427
- CVSS 4.3 Medium
Improper Neutralization of Input Used for LLM Prompting in GitLab
- GitLab
- CWE-1427
- CVSS 8.4 High
[Eclipse Theia] Indirect Prompt Injection via Auto-Loaded Workspace Prompt Template Files in AI Chat
- CWE-829
- CWE-1427
<1.71.0 · fixed in 1.71.0
GHSA-m973-pr9r-hp2w · CVE-2026-46580 - CVSS 8.4 High
[Eclipse Theia] Indirect Prompt Injection via Adversarial Workspace File and Directory Names in AI Chat
- CWE-829
- CWE-1427
<1.71.0 · fixed in 1.71.0
GHSA-3jww-hxqj-wfq2 · CVE-2026-44688 - CVSS 9.8 CriticalCritical · CVSS 9+
MCP Calculate Server: Prompt Injection to RCE
- 611711Dark mcp_calculate_server
- Code injection
- CWE-1427
- CVSS 7.2 High
nnU-Net: Agentic workflow injection in `.github/workflows/issue-triage.yml` of `MIC-DKFZ/nnUNet`
- MIC-DKFZ nnUNet
- CWE-1427
- CVSS 8.7 High
Multiple vulnerabilities in 1millionbot Millie chatbot
- 1millionbot Millie chat
- CWE-1427
- CVSS 5.3 Medium
Salesforce Agentforce Vibes Extension: Improper Neutralization of Input Used for LLM Prompting vulnerability in Salesforce…
- Salesforce Agentforce Vibes Extension
- CWE-1427
- CVSS 6.5 Medium
Salesforce Agentforce Vibes Extension: Improper Neutralization of Input Used for LLM Prompting vulnerability in Salesforce…
- Salesforce Agentforce Vibes Extension
- CWE-1427
- CVSS 5.3 Medium
Salesforce Mulesoft Anypoint Code Builder: Improper Neutralization of Input Used for LLM Prompting vulnerability in Salesforce…
- Salesforce Mulesoft Anypoint Code Builder
- CWE-1427
- CVSS 6.5 Medium
Salesforce Mulesoft Anypoint Code Builder: Improper Neutralization of Input Used for LLM Prompting vulnerability in Salesforce…
- Salesforce Mulesoft Anypoint Code Builder
- CWE-1427
- CVSS 4.6 Medium
Windsurf Prompt Injection via Filename
- Windsurf
- CWE-1427
- CVSS 6.4 Medium
Improper Neutralization of Input Used for LLM Prompting in GitLab
- GitLab
- CWE-1427