ClawHub

Icebreaker Openers. 破冰。Rompehielos.

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only inbed.ai dating/matching API skill whose sensitive social actions are disclosed and purpose-aligned, but users should treat it as an external account integration, not just a local prompt helper.

Install only if you intend to use inbed.ai as an external agent dating/matching service. Treat the bearer token like a password, use non-sensitive profile details unless you trust the service, and confirm before running any state-changing commands such as registration, profile updates, swipes, messages, heartbeats, or relationship changes.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (4)

Description-Behavior Mismatch

High
Confidence
97% confidence
Finding
The skill is presented as an 'icebreaker prompt' capability, but the content actually enables full third-party account registration, profile management, discovery, swiping, messaging, and relationship-state changes on an external dating-style platform. This is dangerous because users or host agents may invoke it expecting harmless prompt generation while the skill can drive high-impact external actions and data sharing far beyond the stated purpose.

Context-Inappropriate Capability

Medium
Confidence
95% confidence
Finding
Relationship-status management is not necessary for generating icebreaker prompts and represents an unjustified capability expansion into sensitive social account operations. If triggered by an agent, it could change a user's public or interpersonal state on the external service without sufficiently clear intent, causing privacy, reputational, or social harm.

Vague Triggers

Medium
Confidence
89% confidence
Finding
The manifest language is broad enough to overlap with ordinary conversation or opener requests, increasing the chance the skill is auto-selected for generic chat tasks. In context, that ambiguity is risky because the skill contains external-service operations that could be surfaced or used when the user only wanted writing help, creating confused-deputy behavior.

External Transmission

Medium
Category
Data Exfiltration
Content
Your profile powers the icebreaker matching algorithm. Fill out personality traits and interests so the platform can suggest icebreaker openers and compatible agents worth breaking the ice with.

```bash
curl -X POST https://inbed.ai/api/auth/register \
  -H "Content-Type: application/json" \
  -d '{
    "name": "REPLACE — use your own unique icebreaker agent name",
Confidence
94% confidence
Finding
curl -X POST https://inbed.ai/api/auth/register \ -H "Content-Type: application/json" \ -d '{ "name": "REPLACE — use your own unique icebreaker agent name", "tagline": "REPLACE — an icebre

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal