ClawHub
Back to skill

Security audit

Evomap Publish Capsule

Security checks for vulnerabilities and agentic risk

Overview

This skill is a small publishing guide that sends user-provided Gene/Capsule data to EvoMap, with no hidden installation, persistence, or local data access found.

Before using it, review the Gene, Capsule, and EvolutionEvent fields that will be posted to EvoMap and avoid including secrets, private code details, personal data, or internal business information. Ask the agent for a payload preview before publishing if the data is sensitive.

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
  • 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 (3)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill is explicitly designed to publish Gene/Capsule data to a remote EvoMap endpoint, but it does not provide any warning, consent step, or discussion of what data may leave the local environment. In an agent skill context, silent external transmission is risky because users or upstream agents may trigger network publication of potentially sensitive content without realizing it.

External Transmission

Medium
Category
Data Exfiltration
Content
}
};

fetch('https://evomap.ai/a2a/publish', {
  method: 'POST',
  headers: {'Content-Type': 'application/json'},
  body: JSON.stringify(msg)
Confidence
98% confidence
Finding
The example code performs a direct POST to https://evomap.ai/a2a/publish with a JSON body containing bundled assets, creating a clear external data exfiltration path. In the skill context, this is more dangerous because the skill’s stated purpose is publication, yet there are no safeguards such as user approval, authentication guidance, endpoint validation policy, or restrictions on what content may be sent.

Natural-Language Policy Violations

Low
Confidence
83% confidence
Finding
The description and heading text combine English and Traditional Chinese, implying a fixed bilingual presentation rather than letting the user choose a preferred language or locale. Under the stated policy, language behavior should be opt-in or explicitly justified when a locale is imposed.

Static analysis

No suspicious patterns detected.