ClawHub
Back to skill

Security audit

RelayShield MCP Firewall

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed setup guide for routing selected MCP traffic through a local security proxy, but users should understand it changes MCP routing and depends on external RelayShield code.

Install only if you intend to route MCP traffic through RelayShield. Review the external RelayShield repository and its documentation before running it, especially because MCP tool calls and results may include sensitive data, credentials, or PII, and the skill recommends a persistent mcporter home-scope configuration.

A.I.G

Vulnerability Patterns
  • 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
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger list contains broad security-related phrases such as 'mcp firewall', 'tool screening', and 'poisoned tools' that could cause the skill to activate during ordinary discussions about MCP security rather than only when the user explicitly wants this skill. Unintended activation is risky here because the skill influences network-routing and security tooling decisions, which may alter agent behavior in sensitive contexts.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% 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 106)May include surrounding context.

md
- The proxy never describes a result as safe. Best case is no flags found.
- If TI screening times out, calls fail open (allowed) and the timeout is logged.
- Stdio MCP servers cannot be proxied over HTTP. For stdio servers, run the pre-deployment scanner and rely on OpenClaw's approval hooks.
- More info: https://api.relayshield.net/developers

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The linked developer endpoint is presented as the sole source of more information and uses an English-only domain path without any indication of locale choice or multilingual support. Under the stated policy, skills should not force a specific language or locale unless clearly justified or optional.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.