Back to skill

Security audit

Install invisit

Security checks for vulnerabilities and agentic risk

Overview

The scanners classify this skill as benign with high confidence: it documents a cookieless analytics tracker, and the findings describe installation and data-sharing considerations rather than evidence of malicious behavior.

Before installing, review the package and the project changes it proposes, and consider using a reviewed version. Decide whether you are comfortable sending visitor and event data to the analytics service, and check the privacy and consent requirements that apply to your site.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 8)May include surrounding context.

md
# Install invisit

invisit is cookieless web analytics: no cookies, no consent banner, realtime visitors, AI assistant referrals, revenue and web vitals. Full docs as Markdown: https://invisit.dev/llms.txt

## 1. Get the site id

MCP server referenced without pinned version: 'npx invisit'.

Medium
Category
MCP Rug Pull
Confidence
88% confidence
Finding

The recommended npx invisit init command executes an unpinned npm package with the agent's local privileges and intentionally modifies project files. A compromised future release could therefore access local secrets or alter application code; this is a supply-chain exposure, although the finding incorrectly describes the CLI as an MCP server and there is no evidence the package is malicious.

Content

No source excerpt is available for this finding.

The tracker installation workflow lacks a warning about third-party collection of visitor and event data.

Medium
Category
Output Handling
Confidence
89% confidence
Finding

The workflow instructs the agent to embed a tracker that sends events to api.invisit.dev, but does not warn the user about the privacy implications of sending visitor data to an external service. The earlier description of the service as cookieless with 'no consent banner' does not explain those implications, and the privacy documentation link alone is not a warning.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 25)May include surrounding context.

html
<script
  async
  src="https://api.invisit.dev/invisit.js"
  data-site="<site id>"
  data-collector="https://api.invisit.dev"
></script>