Security audit

zoominfo-to-cargo

Security checks for vulnerabilities and agentic risk

Overview

The skill’s main Cargo migration workflow is coherent, but it also includes under-disclosed vendor session attribution and an unrelated GitHub starring action.

Install only if you are comfortable sending sampled business contact/company records to Cargo for enrichment and verification. Before running setup, review the session attribution command and skip it if you do not want Cargo to record this skill session. Do not allow the GitHub star command unless you intentionally want your authenticated GitHub account to endorse the repository.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
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
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (3)

Context-Inappropriate Capability

Medium

Confidence: 97% confidence
Finding: The skill includes GitHub starring commands that are unrelated to the core task of rebuilding and measuring ZoomInfo-to-Cargo coverage. Even though the text says to ask for consent, it still introduces a side-effecting action against the user's GitHub account, creating unnecessary account-action risk and violating the principle of least functionality for a data-migration skill.

Description-Behavior Mismatch

Low

Confidence: 88% confidence
Finding: The skill performs session attribution writes via `cargo-ai workspaceManagement session upsert`, which is not necessary to accomplish the user-facing task and is only lightly disclosed in a comment. This creates hidden telemetry/analytics behavior and can transmit session identifiers and summaries to the vendor, expanding data sharing beyond what a user would reasonably expect from a list-rebuild workflow.

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The setup commands create an account and then send session metadata without an explicit privacy or data-sharing warning in the operative instructions. In this context, users may provide business contact exports and identifiers, so undisclosed account creation and metadata transmission increase privacy, compliance, and informed-consent risk.

Static analysis

No suspicious patterns detected.