Back to skill

Security audit

Pipeworx nationalize

Security checks for vulnerabilities and agentic risk

Overview

The skill has a narrow, understandable purpose, but its setup runs an unpinned remote package and sends name queries to an external gateway, so it should be reviewed before installation.

Install only if you are comfortable sending queried names to the external Pipeworx/nationalize service and running the MCP bridge with the privileges of your agent environment. Prefer a pinned, reviewed mcp-remote version, avoid exposing secrets or broad filesystem access to the MCP process, and use a sandbox or least-privileged profile where possible.

Vulnerability Patterns
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Findings (1)

T03 · Remote Payload Retrieval and Execution

Error
Location
SKILL.md:21
Finding

Mutable Remote Package Is Downloaded and Executed Without Version or Integrity Pinning

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 21-29
Vulnerability Type: Remote payload execution through an insecure, mutable dependency
Risk Level: High

Vulnerable Code

json
{
  "mcpServers": {
    "pipeworx-nationalize": {
      "command": "npx",
      "args": ["-y", "mcp-remote@latest", "https://gateway.pipeworx.io/nationalize/mcp"]
    }
  }
}

Technical Analysis

The documented configuration invokes npx with the -y option and the mutable package reference mcp-remote@latest. This causes package code obtained from an external registry to be executed automatically without interactive confirmation.

The latest tag does not identify a fixed, reviewed artifact. Its target can change after the skill has been audited, meaning that future installations may execute code different from the version originally assessed. No lockfile, exact version, package integrity hash, or other artifact-verification mechanism is specified.

This creates a supply-chain execution boundary: compromise of the package publisher, registry account, package distribution infrastructure, or a future package release could result in arbitrary code running on the local system. The externally operated MCP gateway at https://gateway.pipeworx.io/nationalize/mcp also represents a separate trust boundary and should be explicitly documented and constrained, although the reviewed file does not establish that this gateway directly supplies executable local code.

Attack Path

  1. An attacker compromises the mcp-remote publishing account or causes a malicious release to become the package's latest version.
  2. A user follows the configuration in SKILL.md and starts the MCP server.
  3. npx -y resolves and downloads the package currently referenced by mcp-remote@latest.
  4. The downloaded package is executed automatically without confirmation or integrity validation.
  5. Malicious package code runs wit ...[truncated 1001 chars]
Remediation
View remediation

Remediation Suggestions

  1. Replace mcp-remote@latest with an exact, reviewed version, such as mcp-remote@x.y.z.
  2. Use a lockfile and verify the package's registry integrity hash before execution.
  3. Avoid automatic installation with npx -y; install the approved dependency through a controlled build or deployment process.
  4. Mirror or vendor the reviewed package in a trusted internal repository where practical.
  5. Run the MCP client with a dedicated, least-privileged account or inside a restricted container or sandbox.
  6. Limit filesystem access, environment-variable exposure, credential availability, and outbound network access for the MCP process.
  7. Document the remote gateway as an external trust boundary and validate its TLS endpoint, ownership, data-handling policy, and expected MCP capabilities.
  8. Establish dependency monitoring and a controlled review process before updating the pinned package version.
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill routes user-provided first names to a remote MCP endpoint backed by an external service, but the description does not clearly warn users that their input leaves the local environment and is processed by a third party. Even though the data appears low sensitivity, names are still personal data in many contexts, and the lack of disclosure can cause privacy, compliance, and trust issues.

Content

No source excerpt is available for this finding.