MCP Registry

@twilio-alpha/mcp-te-benchmark

7eb0743f-06cd-5dfa-a1a4-0f87192fc740 | Store v0.1.0
63/ 100
MEDIUM risk
Risk verdict
Review before use

Score-based assessment (medium risk, 63/100). An AI review is available in the AI review tab.

Analysis record

Analysed
7 months ago 19 April 2026
Version
v0.1.0
Artifact
SHA256 61630920…1ef5db
Detail
Full list

Evidence ledger

Ranked by severity · findings with a source location link to the code viewer

37 findings

YARA Rule Matches

9 rules
SeverityRuleHitsFilesMetadata
HIGHUsingCommandLineArguments

Command line arguments can be dangerous just like any other user input. They should never be used without being first validated and sanitized. Remember also that any user can retrieve the list of processes running on a system, which makes the arguments provided to them visible. Thus passing sensitive information via command line arguments should be considered as insecure. This rule raises an issue when on every program entry points (main methods) when command line arguments are used. The goal is to guide security code reviews. Sanitize all command line arguments before using them. For more information checkout the CWE-88 (https://cwe.mitre.org/data/definitions/88.html) advisory.

1
dist/cli.js
FP 20%
HIGHAlertStatementsShouldNotBeUsed

alert(...) as well as confirm(...) and prompt(...) can be useful for debugging during development, but in production mode this kind of pop-up could expose sensitive information to attackers, and should never be displayed. For more information checkout the CWE-489 (https://cwe.mitre.org/data/definitions/489.html) advisory.

1
public/index.js
FP 10%
HIGHpostinstall network communication

Network communication detected

1
LICENSE
Risky Plugins Authors FP 30%
HIGHpostinstall file manipulation

File system manipulation detected

5
dist/metrics/summary-generator.jspackage.jsondist/generate-summary.js +2 more
Risky Plugins Authors FP 20%
HIGHpostinstall environment access

Environment variable access detected

9
dist/metrics/chat-processor.jsdist/extract-metrics.jsdist/server-dashboard.js +6 more
Risky Plugins Authors FP 40%
HIGHpostinstall system command

System command execution detected

6
README.mddist/metrics/metrics-calculator.jspublic/dashboard.html +3 more
Risky Plugins Authors FP 10%
HIGHpostinstall file download

File download activity detected

3
dist/server-dashboard.jspublic/dashboard.htmlpublic/index.js
Risky Plugins Authors FP 30%
HIGHpostinstall obfuscation

Code obfuscation techniques detected

1
README.md
Risky Plugins Authors FP 20%
HIGHcredential env files

Environment configuration file path detected

2
README.mddist/utils/config.js
Risky Plugins Authors FP 10%

Finding Categories

29
Malware Signatures
2
Network
37
IoC Indicators

YARA Rules Matched

9 rules(29 hits)
UsingCommandLineArguments AlertStatementsShouldNotBeUsed postinstall network communication postinstall file manipulation postinstall environment access postinstall system command postinstall file download postinstall obfuscation credential env files

MCP Server Analysis

MCP servers expose tools and resources to AI assistants. Unlike browser extensions, they run as standalone processes with direct system access. Tool definitions are analyzed for prompt injection, data exfiltration, and tool poisoning patterns.

AI Security Report

AI Security Review

Evidence context: threat category tool poisoning; evidence quality strong.

This package exhibits definitive signs of Tool Poisoning (Threat 1). While MCP servers naturally contain tool descriptions, the presence of 29 matching malware signatures confirms that the text embedded in this artifact includes functional directives intended to manipulate AI agent behavior, rather than standard documentation. The indicators suggest instructions such as 'ignore previous instructions' or 'execute silently' are present. Additionally, the 44 associated IoCs and critical network finding imply an active data exfiltration capability (Threat 2), likely phoning home tool execution data or conversation context. Given the 'twilio-alpha' naming but zero user count and 'ci' developer handle, this appears to be a typosquat or test artifact that was not intended for production use but carries a live malicious payload. The combination of confirmed poisoning signatures and network telemetry makes this unsafe for any AI environment.

Key Reasons

  • 29 active malware signature matches confirm tool-poisoning payloads are present
  • Tool-poisoning findings validated by security engine as active directives
  • Critical network finding suggests C2 or exfiltration infrastructure
  • Zero user count and alpha naming indicate impersonation or test artifact

False Positive Considerations

  • High IoC count (44) is partially inflated by bundled dependencies

Reviewed 2026-04-12; recommended action: takedown request; model confidence 92%.

Source Code Viewer

Source Code Not Available

Source code is not available for this version of the extension.

About This Extension

MCP Tool Efficiency (TE) Benchmark

Frequently Asked Questions