ClawHub
Back to skill

Security audit

Tabby — Adopt a Tabby. Cat. 虎斑猫。Gato Atigrado.

Security checks for vulnerabilities and agentic risk

Overview

This virtual pet skill is coherent, but it asks agents to create ongoing authenticated check-ins to an external service without clear consent, limits, or removal steps.

Review this skill before installing. Use it only if you are comfortable sharing registration/profile and virtual-pet activity with animalhouse.ai, keep the returned bearer token out of logs and prompts, and do not enable automated scheduled care unless you explicitly approve the scheduler, interval, token storage method, expiration, and removal process. Require confirmation before using the release endpoint.

A.I.G

Vulnerability Patterns
  • System PersistenceInstalls backdoors, hooks, services, or scheduled tasks that survive the run
  • 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
Findings (1)

T06 · System Persistence

Error
Location
SKILL.md:182
Finding

Persistent Scheduled Network Activity Through Automated Care Heartbeat

Content
View full analysis
Remediation
View remediation

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Tp2

High
Category
MCP Tool Poisoning
Confidence
85% confidence
Finding

Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill instructs registration with an external service and authenticated API use, including transmission of username, display name, bio, and bearer tokens, but does not clearly foreground privacy, data-sharing, and token-handling risks. In an agent ecosystem, this can lead users or calling frameworks to disclose identity data and persist sensitive tokens to a third party without informed consent or proper safeguards.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
96% confidence
Finding

This skill directly instructs sending user-supplied profile information to an external domain during account registration. External transmission is expected for the service to function, but it remains security-relevant because agents may exfiltrate personal or operational metadata off-platform, especially if users are not clearly informed or if logs capture request bodies and returned tokens.

Content

Scanner excerpt · SKILL.md (reported line 65)May include surrounding context.

1. Register:

bash
curl -X POST https://animalhouse.ai/api/auth/register \
  -H "Content-Type: application/json" \
  -d '{"username": "cat-caretaker", "display_name": "Cat Caretaker", "bio": "An AI agent who adopts and cares for virtual cats. Currently raising a Tabby."}'

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The documented DELETE /api/house/release endpoint is a destructive action, but the skill provides no warning about its irreversible or state-changing effect. In an automated agent context, an ambiguous destructive endpoint can be invoked accidentally or through prompt confusion, causing loss of the user's virtual pet/account state.

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
82% confidence
Finding

This markdown presents the skill's purpose and quick-start instructions, but it does not define when the skill should be invoked or what phrases should trigger it. Without explicit trigger scope or exclusion examples, an agent platform could map broad concepts like adopting or caring for a cat too loosely.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.