认知自我进化 · 智能体能力进化引擎

• Outcome: Pass
• Latest audit: May 31, 2026, 10:52 AM
• Version:
  1.0.0

Overview

This markdown-only skill is purpose-aligned for analyzing agent failure logs and generating local improvement notes, but users should scope its log access carefully.

Install only if you are comfortable letting the agent read the LOG_DIR you configure. Use a narrow log directory, avoid logs containing secrets, and prefer conservative or observe_only mode if you want to review generated patches before they affect future behavior.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: The skill explicitly instructs the agent to read execution logs from LOG_DIR and extract task IDs, tool call chains, error details, and user feedback, but it does not present a clear privacy warning, consent requirement, or data-handling limitation. This can expose sensitive operational data or personal information contained in logs, especially because the skill also recommends continuous automatic analysis after each failed task.

VirusTotal

64/64 vendors flagged this skill as clean.