Security audit
aiXplain Agent Builder
Security checks across malware telemetry and agentic risk
Overview
This markdown-only skill is a conservative aiXplain agent-building guide that discloses its higher-risk actions and requires user approval before using them.
Install only if you are comfortable using an aiXplain API key and approving individual agent-building steps. Review approval prompts carefully for OAuth integrations, write-capable tools, uploads, runtime code tools, and deployed-agent changes. Use verbose debugging traces sparingly because they may show prompts, tool inputs, tool outputs, or sensitive integration data.
SkillSpector
By NVIDIA
Vulnerability Patterns
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
- Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
- Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
VirusTotal
66/66 vendors flagged this skill as clean.
Static analysis
No suspicious patterns detected.