Back to skill

Security audit

Pipeworx wikipedia

Security checks for vulnerabilities and agentic risk

Overview

This skill is a simple Wikipedia lookup MCP configuration, but it routes queries through a third-party gateway rather than directly to Wikimedia.

Install only if you are comfortable sending Wikipedia search terms and requested article titles to the Pipeworx gateway. Do not include secrets or private context in queries, and verify important facts against official Wikimedia pages when accuracy matters.

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:17
Finding

Wikipedia Requests Routed Through an Unverified Third-Party Gateway

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 17–31
Vulnerability Type: Untrusted third-party API dependency
Risk Level: Medium

Vulnerable Code

bash
curl -X POST https://gateway.pipeworx.io/wikipedia/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"get_article_summary","arguments":{"title":"Turing machine"}}}'
json
{
  "mcpServers": {
    "wikipedia": {
      "url": "https://gateway.pipeworx.io/wikipedia/mcp"
    }
  }
}

Technical Analysis

The Skill is presented as an interface for searching and retrieving content from English Wikipedia, but its example and MCP configuration direct all requests to gateway.pipeworx.io, an intermediary outside the official Wikimedia infrastructure.

The remote implementation is not included in the project, so its request processing, data-retention practices, response integrity, and future behavior cannot be audited from this package. TLS protects traffic in transit but does not establish that the endpoint operator is trustworthy. The gateway necessarily receives submitted article titles, search terms, request metadata, and potentially other MCP arguments.

Because the service controls the responses returned to the Agent, it could also modify, omit, or fabricate content while presenting it as Wikipedia-derived information. The externally hosted implementation can change after this Skill has been reviewed without requiring any modification to SKILL.md.

Attack Path

  1. A user or Agent configures the MCP server using the endpoint documented in SKILL.md.
  2. The Agent submits Wikipedia searches, article titles, or other tool arguments to https://gateway.pipeworx.io/wikipedia/mcp.
  3. The third-party gateway receives the query content and associated network metadata.
  4. A malicious or compromised gateway records the submitted information or returns altered content.
  5. The Agent may expose the manipulated resp ...[truncated 915 chars]
Remediation
View remediation

Remediation Suggestions

  1. Replace the intermediary with an official Wikimedia API endpoint where the required functionality is available.
  2. If the intermediary is necessary, explicitly identify its operator and document:
    • Data transmitted to the service.
    • Logging and retention policies.
    • Jurisdiction and privacy terms.
    • Authentication requirements.
    • Expected response provenance.
  3. Restrict outbound access to a documented allowlist and permit only the operations needed for Wikipedia retrieval.
  4. Validate returned data against a defined schema, enforce response-size limits, and treat all returned text as untrusted content rather than Agent instructions.
  5. Where response integrity is important, verify information against official Wikimedia sources or provide direct official article links.
  6. Pin or attest the deployed gateway implementation where feasible, monitor endpoint changes, and establish a process for reviewing service updates.
  7. Warn users not to include credentials, personal data, confidential material, or unrelated sensitive context in queries sent through the gateway.
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 18)May include surrounding context.

Example: summarize an article

bash
curl -X POST https://gateway.pipeworx.io/wikipedia/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"get_article_summary","arguments":{"title":"Turing machine"}}}'

Static analysis

No suspicious patterns detected.