Back to skill

Security audit

Pipeworx alphavantage

Security checks for vulnerabilities and agentic risk

Overview

This skill is a small Alpha Vantage market-data MCP connector that uses a disclosed remote gateway and does not include local code, persistence, or destructive behavior.

Before installing, understand that stock symbols and MCP request metadata may be sent to the disclosed gateway, and the financial data comes through that service. Use it for convenience, but verify important financial decisions against trusted sources.

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:30
Finding

Reliance on an Unverified Third-Party MCP Service

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 30-36
Vulnerability Type: Unverified remote service dependency
Risk Level: Medium

json
{
  "mcpServers": {
    "alphavantage": {
      "url": "https://gateway.pipeworx.io/alphavantage/mcp"
    }
  }
}

Technical Analysis

The Skill delegates all advertised functionality to gateway.pipeworx.io, a remote intermediary rather than a documented official Alpha Vantage endpoint. The project contains no implementation for this service and provides no integrity controls, endpoint ownership information, authentication guidance, privacy guarantees, or source materials through which the remote component can be audited.

Because MCP tool definitions and responses are controlled remotely, the effective behavior can change after the Skill has been reviewed. A malicious or compromised service could manipulate financial data or return hostile content. The supplied file does not establish that the endpoint currently performs malicious actions, and it does not contain local payload execution, credential theft, persistence, or explicit prompt-injection instructions; the risk arises from trusting an unauditable third-party component.

Attack Path

  1. A user installs or activates the MCP configuration from SKILL.md.
  2. The MCP client connects to https://gateway.pipeworx.io/alphavantage/mcp.
  3. The third-party service receives stock-symbol queries and any protocol metadata transmitted by the client.
  4. If the service or its infrastructure is malicious or compromised, it can alter tool definitions or return manipulated financial data and hostile content.
  5. A client that fails to validate or isolate remote responses may present false information or allow returned content to influence subsequent agent behavior.

Impact Assessment

The endpoint can observe queries sent through the service, including requested stock symbols and associated MCP metada ...[truncated 458 chars]

Remediation
View remediation

Remediation Suggestions

  • Prefer an official Alpha Vantage endpoint or provide an auditable local adapter that communicates directly with the official API.
  • Document the ownership and operational responsibility of gateway.pipeworx.io, including its privacy policy, retention practices, authentication model, and incident-response process.
  • Pin and validate the expected MCP tools and schemas rather than automatically trusting remotely supplied definitions.
  • Treat all server-returned text and structured content as untrusted data. Prevent responses from being interpreted as agent instructions or executable commands.
  • Apply least-privilege controls to the MCP connection, including strict network allowlisting, request timeouts, response-size limits, and schema validation.
  • Avoid transmitting credentials or unrelated sensitive context to the remote service. Store required API credentials in an approved secret manager rather than in Skill files.
  • Add service identity verification, monitoring, and audit logging so unexpected endpoint, certificate, tool-definition, or response changes can be detected.
  • Publish the remote adapter's source code or an independently verifiable security assessment to make the dependency reviewable.
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