ClawHub
Back to skill

Security audit

Graceful Boundaries

Security checks across malware telemetry and agentic risk

Overview

This skill transparently audits or implements rate-limit communication behavior and does not show hidden install, credential, persistence, or destructive behavior.

Install if you want an agent to assess or implement Graceful Boundaries behavior. Run audits only against services you own or are authorized to test, avoid pointing the checker at internal or sensitive network targets unless intended, and review any builder-mode code changes before committing or deploying.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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
Findings (1)

Lp3

Medium
Category
MCP Least Privilege
Confidence
94% confidence
Finding
The skill explicitly instructs the agent to make outbound HTTP requests ('Fetch the limits discovery endpoint directly' and 'fetch one of them') and even suggests using curl, fetch, or another HTTP client, but the metadata declares no permissions for network access. This creates a capability/permission mismatch that can bypass operator expectations, leading to unintended SSRF-style access to internal resources, access to sensitive local network endpoints, or policy violations if the runtime does not strictly gate egress.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.