ClawHub
Back to skill

Security audit

PostHog

Security checks across malware telemetry and agentic risk

Overview

This PostHog integration is purpose-aligned and clearly warns about authentication, user approval for writes, and credential handling, with one minor example-sanitization concern.

Before installing, be aware that this skill can access PostHog analytics data and can modify PostHog resources if you approve write calls. Prefer OAuth, choose the narrowest PostHog scopes available, confirm account and connection IDs before changes, and avoid displaying or storing token fields from API responses.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Intent-Code Divergence

Medium
Confidence
93% confidence
Finding
The skill explicitly warns that provider-issued tokens and credentials must never be printed or persisted, yet its example response includes an `api_token` field with a token-like value (`phc_XXX`). Even if redacted, normalizing the display of credential-bearing fields in examples increases the chance that an agent will echo real secrets from live responses back to the user or logs, contradicting the skill’s own handling rules.

VirusTotal

57/57 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.