Wrike

Security checks across malware telemetry and agentic risk

Overview

This Wrike skill is a mostly transparent API guide, but it includes sensitive admin, audit-log, and bulk export capabilities that are broader than its project-management framing makes clear.

Install only if you intend to let the agent use broad Wrike API access through Maton, potentially including admin reads and bulk exports if the connected account permits them. Use a least-privileged Wrike account, keep MATON_API_KEY private, specify the intended Maton connection when multiple accounts exist, and require explicit confirmation before any write, delete, audit-log, user-directory, attachment-download, or data-export request.

SkillSpector

By NVIDIA

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (5)

Description-Behavior Mismatch

Medium

Confidence: 94% confidence
Finding: The skill claims access is limited to ordinary work-management resources, but the documented surface also includes administrative functions such as user management, audit logs, invitations, access roles, and data export. This mismatch can mislead users or downstream agents into granting or invoking broader access than they intended.

Context-Inappropriate Capability

High

Confidence: 96% confidence
Finding: Bulk data export enables large-scale extraction of organizational data well beyond normal task management. In a general Wrike productivity skill, this materially increases the risk of mass data exfiltration, especially because the manifest does not justify or prominently warn about this capability.

Context-Inappropriate Capability

Medium

Confidence: 89% confidence
Finding: Audit log access exposes sensitive administrative telemetry such as login events, user emails, IP addresses, and operational history. That is a broader monitoring capability than the stated project-management purpose and creates privacy and surveillance risks if invoked without explicit justification.

Context-Inappropriate Capability

Medium

Confidence: 91% confidence
Finding: User, invitation, and access-role administration affect account governance and membership, not just project work. Bundling these capabilities into a broad Wrike skill increases the chance of unauthorized or accidental identity/admin changes under an apparently routine invocation path.

Missing User Warnings

Medium

Confidence: 87% confidence
Finding: The documentation includes audit and export features that can expose sensitive organizational data, but it does not provide proportional user-facing warnings about privacy, surveillance, or mass-export consequences. This omission increases the chance that an agent or user will treat high-risk operations as routine.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal