Memorine

Security checks across malware telemetry and agentic risk

Overview

Memorine is a coherent local agent-memory skill, but users should treat its stored and shared memories as persistent sensitive data.

Install only if you are comfortable with agents writing long-lived memory to a local SQLite database. Avoid storing secrets, credentials, regulated data, or private customer information unless you have clear filesystem protections, backup handling, deletion procedures, and separate databases or namespaces for users, teams, or tenants that should not share memory.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
System Prompt LeakageDirect Leakage, Indirect Extraction, Tool-Based Exfiltration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (3)

Missing User Warnings

Medium

Confidence: 86% confidence
Finding: The README promotes cross-agent sharing and team-wide knowledge without clearly warning that stored facts may be visible to other agents or users sharing the same database or namespace context. In an agent-memory product, this can lead to unintended disclosure of sensitive operational details, credentials, infrastructure metadata, or user data if developers assume memory is private by default.

Missing User Warnings

Medium

Confidence: 89% confidence
Finding: The README says the generated cognitive profile can be dropped into a system prompt, but does not warn that the profile may contain sensitive facts, events, or team knowledge previously stored in memory. Putting such content into prompts can expand exposure to downstream models, logs, telemetry systems, or prompt inspection surfaces, causing accidental data leakage.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill exposes persistent storage of facts, events, procedures, and cross-agent sharing, but the description section does not prominently warn users that sensitive agent data is retained in a local SQLite database and may be shared across agents. This can lead operators to enable the skill without understanding the privacy, retention, and data-governance implications, increasing the risk of unintended storage or disclosure of secrets, personal data, or operational history.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal