Back to skill

Security audit

Pipeworx crossref

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent Crossref lookup integration, but its setup runs an unpinned npm package at install/runtime, which deserves review before use.

Install only if you are comfortable running the MCP bridge in a restricted environment. Prefer pinning `mcp-remote` to a reviewed exact version, and do not send sensitive or confidential queries through the hosted gateway.

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

Error
Location
SKILL.md:43
Finding

Automatic Execution of an Unpinned npm Dependency

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 43–50
Vulnerability Type: Unpinned third-party dependency executed at runtime
Risk Level: High

Vulnerable Code

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

Technical Analysis

The recommended MCP configuration invokes npx with both the automatic-confirmation option (-y) and the mutable package selector mcp-remote@latest. Consequently, the MCP client can download and execute whichever package release the npm registry identifies as latest at invocation time.

Because neither an exact audited version nor a package integrity hash is specified, the effective executable code can change after this Skill has been reviewed. A compromise of the npm package, its publisher account, or its dependency chain could therefore introduce arbitrary code into subsequent executions without requiring any modification to SKILL.md.

This finding does not establish that the current package is malicious. The vulnerability is the configuration's implicit trust in mutable, remotely retrieved executable code.

Attack Path

  1. A user copies the documented MCP configuration into a client.
  2. An attacker compromises the mcp-remote npm package, its publisher account, or a dependency used by a future release.
  3. The attacker publishes a modified release and causes it to resolve through the latest tag.
  4. The MCP client starts the configured server and invokes npx -y mcp-remote@latest.
  5. npx retrieves the attacker-controlled package without an interactive approval prompt.
  6. npm package lifecycle logic or the package entry point executes with the privileges of the user running the MCP client.

Impact Assessment

Successful exploitation could execute arbitrary code under the MCP client user's account. Depe ...[truncated 371 chars]

Remediation
View remediation

Remediation Suggestions

  1. Replace mcp-remote@latest with an exact, reviewed version such as mcp-remote@X.Y.Z; do not use mutable tags or version ranges.
  2. Install the dependency during a controlled deployment step rather than downloading and executing it automatically whenever the MCP server starts.
  3. Commit and enforce a lockfile containing integrity metadata, and verify the package and its transitive dependency hashes before installation.
  4. Review the package's source repository, publisher identity, release provenance, lifecycle scripts, and dependency tree before approving a version.
  5. Disable unnecessary npm lifecycle scripts where operationally possible.
  6. Run the MCP process in a restricted environment with minimal filesystem access, no unnecessary credentials, and constrained network permissions.
  7. Establish a deliberate update process in which new versions are security-reviewed and tested before the pinned version is changed.
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/crossref
---