Stability Ai
Security checks across malware telemetry and agentic risk
Overview
This skill is a disclosed Stability AI integration that uses Membrane for authentication and API access, with expected privacy and credential considerations but no evidence of hidden or malicious behavior.
Install only if you trust Membrane's CLI and delegated credential handling. Use it for explicit Stability AI tasks, avoid sending sensitive prompts or media unless necessary, and confirm requests that may consume credits or affect account state.
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
65/65 vendors flagged this skill as clean.