ClawHub
Back to skill

Security audit

Uplo Insurance

Security checks for vulnerabilities and agentic risk

Overview

This skill fits its insurance knowledge-management purpose, but it needs review because it gives an unpinned external MCP server broad access to sensitive insurance context and export-style tools.

Install only after reviewing the UPLO tenant, token scope, and MCP package provenance. Use a dedicated least-privilege token, pin or verify the MCP server package, restrict accessible packs and classification tiers, and require human approval before full organization exports or any action that changes shared knowledge-base state.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Missing User Warnings

Medium
Confidence
84% confidence
Finding
The README advertises an `export_org_context` capability that produces a full organizational context snapshot, but it provides no warning about privacy, least-privilege access, or sensitive data handling. In an insurance-focused skill, organizational context is likely to include highly sensitive policy, claims, underwriting, and possibly regulated personal data, so normalizing unrestricted export increases the risk of over-collection, unauthorized disclosure, and misuse.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The manifest configures an external HTTP MCP endpoint and exposes search and export-style capabilities over insurance data, but provides no user-facing warning or consent boundary about transmitting potentially sensitive policy, claims, underwriting, or actuarial information to a third-party service. In an insurance context, this is especially risky because users may send regulated personal, financial, or health-adjacent claim data off-platform without realizing it, creating confidentiality, compliance, and data-handling exposure.

Static analysis

No suspicious patterns detected.