Retrospective Agent
Security checks across malware telemetry and agentic risk
Overview
This is a markdown-only retrospective skill that stores visible local lesson notes and includes clear limits against hidden memory, secrets, persona changes, and autonomous behavior.
Install this if you want the agent to keep auditable retrospective notes in the workspace. Review entries before treating them as durable behavior rules, and avoid using the skill to store secrets, sensitive personal information, or hidden preferences.
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.