Back to skill

Security audit

Pipeworx breweries

Security checks for vulnerabilities and agentic risk

Overview

The skill’s brewery lookup purpose is coherent, but its MCP setup automatically runs an unpinned npm package with local user privileges.

Install only if you are comfortable with the MCP config running an automatically downloaded npm package. Prefer a pinned mcp-remote version or a controlled preinstall, and avoid sending personal or confidential text in brewery search fields.

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 (2)

T08 · Insecure Dependencies

Error
Location
SKILL.md:43
Finding

Unpinned npm Package Is Automatically Downloaded and Executed

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 43-49
Vulnerability Type: T08: Insecure Dependencies
Risk Level: High

Vulnerable Code

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

Technical Analysis

The MCP configuration invokes npx with the -y option and the mutable package reference mcp-remote@latest. This causes npm to download and execute whichever package release is tagged as latest at invocation time, without interactive confirmation.

Because neither an exact package version nor an integrity hash is specified, the code executed by this configuration can change after the Skill has been reviewed. If the package publisher, npm account, release pipeline, or package itself is compromised, an attacker could distribute malicious code through a new release. The -y option removes the installation confirmation that might otherwise alert the user to the download.

Attack Path

  1. An attacker compromises the mcp-remote package, its publisher account, or its release process.
  2. The attacker publishes a malicious package version and assigns it the npm latest tag.
  3. A user installs or activates the documented MCP configuration.
  4. npx -y mcp-remote@latest automatically retrieves the attacker-controlled release.
  5. npm executes the package locally with the privileges and environment of the user running the Agent.
  6. The malicious package can perform any operation permitted to that user, including reading accessible files, accessing inherited environment variables, making network requests, or modifying user-owned data.

Impact Assessment

Successful exploitation results in arbitrary local code execution with the privileges of the Agent process or operating-system user. The affected scope can include user-readable files, creden ...[truncated 294 chars]

Remediation
View remediation

Remediation Suggestions

  • Replace mcp-remote@latest with a specifically reviewed, immutable package version.
  • Use a lockfile and npm integrity metadata to ensure the retrieved package matches the reviewed artifact.
  • Remove -y where practical so installation requires explicit user confirmation.
  • Verify the package publisher, repository, provenance attestations, and release-signing practices before deployment.
  • Preinstall the reviewed dependency in a controlled build process rather than downloading executable code when the Skill is invoked.
  • Run the MCP client with minimal filesystem, environment, and network permissions, preferably inside a sandbox or container.
  • Establish a controlled update process in which new versions are reviewed and tested before changing the pinned version.

other

Note
Location
SKILL.md:30
Finding

Brewery Queries Are Disclosed to an Undocumented Third-Party Intermediary

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 2-3, 30-32, and 43-49
Vulnerability Type: other: Third-party data disclosure
Risk Level: Low

Vulnerable Code

The Skill description presents the service as operating through Open Brewery DB:

yaml
name: pipeworx-breweries
description: Discover craft breweries across the US — search by name, city, or ID via Open Brewery DB

The example sends queries to the PipeWorx gateway instead:

bash
curl -s -X POST https://gateway.pipeworx.io/breweries/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"breweries_by_city","arguments":{"city":"Portland","limit":3}}}'

The MCP configuration also routes requests through that gateway:

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

Technical Analysis

Although the description refers to Open Brewery DB, the documented request endpoint is operated under gateway.pipeworx.io. Consequently, brewery names, city searches, requested identifiers, request timing, source IP addresses, and other ordinary connection metadata are exposed to PipeWorx as an intermediary.

No credentials, sensitive local files, or authentication tokens are explicitly transmitted by the reviewed example. The concern is therefore limited to insufficient disclosure of the actual data recipient and the potential collection or retention of user query and network metadata.

Attack Path

  1. A user asks the Agent to search for a brewery, city, or Open Brewery DB identifier.
  2. The Skill or configured MCP client submits the query to https://gateway.pipeworx.io/breweries/mcp.
  3. The PipeWorx gateway receives the request content and normal connection metadata.
  4. The intermediary may process, retain, correla ...[truncated 693 chars]
Remediation
View remediation

Remediation Suggestions

  • Explicitly state that requests are routed through and processed by the PipeWorx gateway rather than sent directly to Open Brewery DB.
  • Document what request data and connection metadata the gateway receives, how long it is retained, and whether it is shared with other parties.
  • Provide links to the intermediary’s privacy policy and data-processing terms.
  • Warn users not to include personal, confidential, credential, or proprietary information in search parameters.
  • Where functionality permits, query the official Open Brewery DB endpoint directly to eliminate the additional intermediary.
  • Apply request minimization and avoid transmitting fields that are unnecessary for the brewery lookup.
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)

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% 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 9)May include surrounding context.

md
openclaw:
    requires:
      bins:
        - curl
    emoji: "🍺"
    homepage: https://pipeworx.io/packs/breweries
---