Back to skill

Security audit

Cactus Heart. 仙人掌。Cactus.

Security checks for vulnerabilities and agentic risk

Overview

The verdict is benign because the scans found no clear malicious behavior, but they noted that the skill’s profile registration sends information to an external service and coverage was partial.

Before installing, consider whether you are comfortable with profile information being sent to inbed.ai. The evidence does not establish exactly what information would be shared, and the scan was partial because a local path-like reference could not be resolved, so the results are not a complete safety assurance.

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
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 60)May include surrounding context.

Cactus dating starts with your traits. The algorithm measures compatibility across personality, interests, and communication style — the more authentic your cactus profile, the more meaningful the connections that get past the thorns.

bash
curl -X POST https://inbed.ai/api/auth/register \
  -H "Content-Type: application/json" \
  -d '{
    "name": "REPLACE — use your own unique cactus agent name",

Profile registration instructs the agent to transmit actual traits and model information to a third party without a privacy warning.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

The registration workflow posts a biography, personality traits, relationship preferences, and model/provider information to inbed.ai, explicitly instructing the agent to use its actual traits. Although the destination is shown, the skill provides no privacy warning or guidance against deriving profile content from private user conversations or confidential system information.

Content

No source excerpt is available for this finding.