Security audit

Context Booster

Security checks across malware telemetry and agentic risk

Overview

This context-memory skill appears purpose-aligned, but it stores and reuses broad conversation and environment context across sessions without clear consent, limits, or deletion controls.

Install only if you want durable cross-session memory. Treat stored context as potentially sensitive: avoid secrets, check where memory is written, look for commands to inspect or delete it, and prefer using it in a scoped workspace rather than across broad personal or project data.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
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

Findings (5)

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The FAQ states that key conversation context is automatically saved at the end of each session and can be restored in a new session, but it does not clearly warn users at the point of description that persisted data may include sensitive or personal information. In a context-retention skill, automatic cross-session storage materially increases privacy risk because users may assume the conversation is ephemeral while the system preserves data locally for later retrieval.

Missing User Warnings

Medium

Confidence: 83% confidence
Finding: The README explicitly states that context enhancement can pull from historical conversations, long-term memory, external knowledge bases, and related files, but it does not describe any consent model, access boundaries, redaction rules, or privacy safeguards. In a context-management skill, this omission is security-relevant because it normalizes broad data aggregation and retrieval that could expose sensitive user data, prior chats, or local/project information beyond user expectations.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill explicitly advertises capturing, compressing, retrieving, injecting, and persistently storing conversation context across sessions, but only much later gives a brief note that sensitive information needs special handling. That is insufficient for a design centered on collecting broad user/session data, because it can normalize retention of secrets, personal data, and system state without consent, minimization, retention limits, or access controls.

Missing User Warnings

Medium

Confidence: 97% confidence
Finding: The code example shows capture of both userInput and ecosystemState(), which could contain credentials, personal data, tokens, file contents, or other sensitive runtime information. Presenting this as a normal usage pattern without prominent safeguards encourages over-collection and unsafe persistence of high-value data.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The version history explicitly advertises context capture and cross-session persistence, which implies retention of potentially sensitive user or session data, but it does not include any warning, consent model, retention limit, or privacy safeguards. In a context-enhancement skill, this omission is security-relevant because operators may enable persistent memory without understanding the privacy and data-handling consequences.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.