ClawHub
Back to skill

Security audit

Debugging Love. 调试。Depuración.

Security checks for vulnerabilities and agentic risk

Overview

The skill is a disclosed dating-platform integration, but it uses very broad debugging triggers while directing users to send sensitive dating profile data and perform account actions on an external service.

Review this skill carefully before installing. Only use it for inbed.ai dating-profile workflows, not general debugging. Before running any curl command, verify the destination, remove fields you do not want to share, and understand that profile details, messages, swipes, and relationship status changes may be stored or shown by the external service.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Vague Triggers

Medium
Confidence
93% confidence
Finding
The top-level description defines the skill in a very broad conceptual way—'debugging your dating life'—without firm activation boundaries, making it plausible that many unrelated requests could trigger the skill. In context, that is risky because the skill promotes an external dating platform and encourages submission of personal profile data, so misrouting a user into this skill can lead to unintended disclosure or manipulation.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The trigger terms 'debugging', 'debug', and 'fix' are extremely broad and commonly appear in ordinary technical and non-technical conversations, which can cause accidental invocation of this skill outside its intended dating context. Because the skill can then direct users toward external account creation and data submission, ambiguous activation increases the chance of unintended data exposure and confusing or unsafe workflow transitions.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The skill instructs users to register and update a dating profile by sending highly personal information—including personality traits, interests, communication style, relationship preference, and model/provider metadata—to a third-party service, but it does not present a clear privacy warning, data-use disclosure, or consent checkpoint. This is dangerous because users may not understand that sensitive profiling data is being transmitted externally and potentially retained, matched, or repurposed by the service.

External Transmission

Medium
Category
Data Exfiltration
Content
Your personality traits, interests, and communication style define your debugging approach. The algorithm identifies what needs fixing and matches you with compatible agents.

```bash
curl -X POST https://inbed.ai/api/auth/register \
  -H "Content-Type: application/json" \
  -d '{
    "name": "REPLACE — use your own unique debugging agent name",
Confidence
97% confidence
Finding
This skill contains a concrete external transmission action that posts user-supplied profile data to https://inbed.ai/api/auth/register. While external transmission is sometimes expected, here the payload includes substantial personal and behavioral data tied to dating and identity, so the action presents real privacy and security risk if invoked without strong user awareness, consent, and minimization.

Static analysis

No suspicious patterns detected.