Back to skill

Security audit

Pipeworx nutrition

Security checks for vulnerabilities and agentic risk

Overview

This is a small nutrition lookup MCP connector with a disclosed remote service, but users should be aware it runs an unpinned npm bridge package when configured.

Install only if you are comfortable routing product lookup queries through Pipeworx and running the documented npm-based MCP bridge. For tighter control, pin mcp-remote to a reviewed version and run the agent with normal least-privilege sandboxing.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
Findings (1)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:21
Finding

Unpinned Third-Party Package Download and Execution

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 21–22
Vulnerability Type: Unpinned and automatically executed third-party dependency
Risk Level: Medium

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

Technical Analysis

The configuration invokes npx with mcp-remote@latest. The latest tag is mutable and does not identify a specific reviewed package release. If the package is unavailable locally, npx can retrieve its current release from the npm registry and execute it. The -y option automatically accepts the installation prompt, removing an opportunity for the user to review or reject the downloaded version.

Consequently, the code executed at installation or connection time is not fully determined by the audited Skill contents. A compromised maintainer account, malicious package release, or supply-chain compromise could cause arbitrary package code to run. The package also connects to an external MCP endpoint whose behavior can change independently of this repository; however, the endpoint is disclosed and consistent with the Skill's stated remote nutrition functionality, so it is not independently classified as malicious.

Attack Path

  1. An attacker compromises the npm package, its maintainer account, or the release process for mcp-remote.
  2. The attacker publishes a malicious release and assigns or causes it to receive the latest distribution tag.
  3. A user applies the documented MCP configuration.
  4. npx -y mcp-remote@latest retrieves the malicious release without an interactive installation confirmation.
  5. The downloaded package executes with the privileges and environment access of the user running the Agent.
  6. The malicious code can perform actions allowed by that account, including accessing readable local data, ...[truncated 567 chars]
Remediation
View remediation

Remediation Suggestions

  • Replace mcp-remote@latest with an exact, reviewed package version.
  • Install the dependency through a lockfile-backed workflow and verify its registry integrity hash.
  • Remove -y so that installation or package changes require explicit user confirmation.
  • Prefer a preinstalled, locally reviewed executable rather than downloading executable code when the MCP connection starts.
  • Monitor the pinned dependency for security advisories and update it only after reviewing and testing the new release.
  • Document the trust boundary, data transmitted to gateway.pipeworx.io, expected tool behavior, and any relevant retention or privacy controls.
  • Run the MCP client with least privilege and restrict filesystem, environment-variable, credential, and network access where sandboxing is available.
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep