Draftable

Security checks across malware telemetry and agentic risk

Overview

This Draftable skill appears purpose-aligned, but it needs review because it gives an agent authenticated access to create, export, delete, and directly call APIs for potentially sensitive documents without enough safety guidance.

Install only if you are comfortable giving Membrane-authenticated Draftable access to an agent. Confirm exact comparison IDs before deletion, review endpoint/method/body before any proxy request, avoid sending raw document contents unless necessary, and revoke the Membrane connection when no longer needed.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill documents a destructive 'Delete Comparison' capability without any warning, confirmation requirement, or statement that deletion should only occur with explicit user intent. In an agent context, this increases the risk of accidental or over-broad deletion of user data if the model selects the action from ambiguous instructions.

Missing User Warnings

Medium

Confidence: 89% confidence
Finding: The proxy section encourages direct API requests but omits any warning that document contents, metadata, and potentially sensitive legal or financial information may be transmitted to a third-party service. Because Draftable is specifically used for sensitive documents, missing data-handling guidance materially increases the chance of inadvertent exposure or misuse.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal