Security audit

Limesurvey Openclaw Skill

Security checks across malware telemetry and agentic risk

Overview

This appears to be a legitimate LimeSurvey automation skill, but it can export sensitive survey data, send participant emails, and change or delete survey records with limited built-in safeguards.

Review before installing in any production LimeSurvey environment. Use a dedicated least-privilege service account instead of admin credentials, confirm survey IDs and recipient lists before mutating actions or invitations, keep backups before delete/import/activation workflows, and treat exported responses, participant lists, tokens, and email addresses as sensitive data that should not be printed, logged, or stored casually.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (9)

Intent-Code Divergence

Medium

Confidence: 96% confidence
Finding: The skill explicitly recommends least-privilege service accounts, but the Quick Start and Python examples then instruct users to authenticate as 'admin'. That contradiction normalizes use of overly privileged credentials, increasing the blast radius of any accidental misuse, credential leakage, or compromise of the automation environment.

Missing User Warnings

Medium

Confidence: 83% confidence
Finding: The README prominently advertises destructive operations such as deleting or activating/importing surveys without warning that some actions may be irreversible or operationally disruptive. In an agent skill context, documentation directly shapes agent behavior, so missing safeguards can increase the chance of accidental destructive actions against production survey data.

Missing User Warnings

Medium

Confidence: 87% confidence
Finding: The README describes exporting responses, managing participants, and sending invitations/reminders without any privacy, consent, or sensitive-data handling warning. Because survey responses and participant records often contain personal or regulated data, omission of handling guidance increases the risk of unauthorized export, disclosure, or misuse by users or downstream agents.

Missing User Warnings

Medium

Confidence: 84% confidence
Finding: The skill shows how to export survey responses and write them directly to local files without any warning that responses may contain sensitive personal, health, demographic, or research data. In this context, silent local export can lead to insecure storage, accidental sharing, or retention of regulated data outside approved controls.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: The documented API surface includes multiple destructive operations such as deleting surveys, participants, responses, groups, and questions, but does not warn users about irreversibility or the need for confirmation and backups. In an automation skill, omission of these safeguards increases the risk of accidental mass data loss through operator error or prompt misuse.

Missing User Warnings

Medium

Confidence: 82% confidence
Finding: This section documents bulk response export, including JSON/CSV/PDF formats, without any warning that survey responses often contain sensitive personal or confidential data. In an agent skill context, making export operations easy to invoke without emphasizing data sensitivity increases the likelihood of overbroad extraction, mishandling, or downstream disclosure.

Missing User Warnings

Medium

Confidence: 83% confidence
Finding: The examples include operations such as activating, copying, importing, and deleting survey data without clearly warning that these calls modify live LimeSurvey state. In an automation skill for survey administration, this is contextually expected, but example code can still be pasted into production and cause unintended changes or data loss if users do not recognize the side effects.

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: These participant-management examples process and display personal data including names, emails, tokens, and custom attributes, but they do not include privacy or data-minimization warnings. In a survey-management skill this functionality is legitimate, yet exposing PII in sample outputs and encouraging broad listing/printing increases the risk of accidental disclosure, especially in shared logs, terminals, or notebooks.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The response export examples decode, print, and write survey responses directly to local files without warning that response content may contain highly sensitive personal or confidential information. Because survey responses often include free text, demographic data, or regulated data, these examples materially increase the chance of unauthorized disclosure through local storage, console output, backups, or downstream handling.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.