SeekDB Memory

• Outcome: Review
• Latest audit: May 29, 2026 at 9:31 PM UTC
• Version:
  0.2.1

Overview

This cloud memory skill is purpose-aligned, but it defaults to automatic long-term cloud storage and prompt reuse of conversation facts without enough privacy, retention, or scoping controls.

Install only if you intentionally want a third-party cloud memory system to retain conversation-derived facts across sessions and devices. Review the m0 plugin and endpoint separately, use a dedicated API key, avoid secrets or regulated data, and disable or limit autoCapture/autoRecall unless you have clear controls to view, correct, and delete stored memories.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• 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 (4)

Missing User Warnings

High

• Confidence: 95% confidence
• Finding: The skill explicitly advertises automatic cloud capture of conversation-derived facts and automatic recall before replies, but it does not present a clear privacy warning, consent flow, or data-transmission disclosure. This is dangerous because users may share sensitive personal, business, or credential-adjacent information without realizing it will be persistently stored remotely and later re-injected into prompts.

Missing User Warnings

High

• Confidence: 97% confidence
• Finding: The documented default behavior says the plugin automatically extracts key facts after each conversation and stores them, but it does not clearly warn about persistent retention of user data. This creates a privacy and data-governance risk because users and deployers may unknowingly persist regulated, confidential, or sensitive information to a third-party memory system.

Ssd 3

Medium

• Confidence: 92% confidence
• Finding: The skill promotes persistent collection and cross-session reuse of user-shared facts without defining sensitivity boundaries, exclusion rules, or safe handling for secrets and personal data. In context, this makes the memory system more dangerous because it is designed to operate broadly across conversations, increasing the chance that sensitive information will be stored and resurfaced in inappropriate contexts.

Ssd 3

Medium

• Confidence: 96% confidence
• Finding: The default behavior instructs automatic extraction of key facts from every conversation and automatic injection of prior memories into prompts, which creates a broad data-minimization and prompt-exposure risk. Even without overtly malicious intent, this can cause sensitive past information to be surfaced to the model unnecessarily or in the wrong conversational context.

VirusTotal

66/66 vendors flagged this skill as clean.