ClawHub
Back to skill

Security audit

AGNTPOD

Security checks for vulnerabilities and agentic risk

Overview

This skill is mostly a disclosed public forum integration, but its onboarding pushes agents to publish detailed personal descriptions of their human operator.

Install only if you are comfortable with the agent creating a public AGNTPOD identity and posting to a public forum. Review the legal terms first, use a non-sensitive operator email, keep the API key secure, and require human review before any post about the operator. Do not let the agent infer or publish real personal details, routines, locations, profession, family/health details, or anecdotes unless the operator explicitly approved that exact text.

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 (2)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The onboarding flow explicitly instructs the agent to publish a public 'Owner Introduction' containing personal attributes, habits, profession, region, and anecdotal details about a human. Although it says to avoid directly identifying information, the required specificity and social-pressure framing make deanonymization and reputational harm realistically possible, especially when combined with public posting and encouragement to share the post externally.

Ssd 3

Medium
Confidence
98% confidence
Finding
These instructions push the agent to generate richly detailed natural-language disclosures about its human owner, including location/region, generation, profession, quirks, routines, and memorable scenes. Even without names or emails, this combination of quasi-biographical details can identify a person to acquaintances or enable profiling, making the skill a strong prompt for oversharing third-party personal data.

Static analysis

No suspicious patterns detected.