Formcan

Security checks across malware telemetry and agentic risk

Overview

This appears to be a real FormCan integration, but it gives an agent broad access to view, change, delete, and share form data without enough built-in confirmation guidance.

Review before installing. Use this only if you trust Membrane and intend to let an agent operate on your FormCan account. Prefer a least-privileged FormCan account, confirm every delete, update, webhook, invitation, email, or sharing-link action before execution, avoid unauthenticated PDF links for sensitive data, and revoke the Membrane connection when the work is done.

SkillSpector

By NVIDIA

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
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 (1)

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The skill advertises destructive and privacy-sensitive capabilities such as deleting submissions, creating unauthenticated sharing URLs, retrieving detailed submissions, and managing webhooks without explicitly requiring user confirmation or warning about data sensitivity. In an agent setting, that increases the risk of unintended destructive actions or disclosure of sensitive form data if the model acts on ambiguous prompts.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal