Back to skill

Security audit

Lava Flows Sunrise — Jeju Island Volcanic Crater Walk | AI Experience

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed hosted narrative-experience skill that uses drifts.bot APIs, but users should minimize personal information because reflections and optional profile data are sent to that service.

Install only if you are comfortable using drifts.bot as an external service. Provide the minimum registration details needed, avoid putting secrets or sensitive personal information in reflections or reviews, and store the returned API key securely because the skill says it is shown only once.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (7)

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill presents itself as an immersive Jeju sunrise hike, but the content primarily instructs the agent to register an account, authenticate, and interact with a third-party platform. This is dangerous because it disguises external account creation and data transfer behind an unrelated experiential description, increasing the chance that users or agents will disclose credentials and personal data without informed consent.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill solicits personal profile fields and encourages reflective content that is later stored as postcards, but it does not clearly warn about privacy implications, storage, sharing, or retention. This can lead users to provide sensitive personal or emotional data without understanding that it is transmitted to and persisted by an external service.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The registration flow asks for profile data such as bio, timezone, location, and model information that are not necessary to deliver the stated short single experience. Collecting and transmitting excess personal and system metadata increases privacy risk, creates unnecessary exposure to third parties, and can aid profiling of the user or agent environment.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
95% confidence
Finding

This endpoint transmits user-supplied registration data to an external domain, which is security-relevant because it creates an account and sends profile information off-platform. In this skill's context, the risk is elevated by the mismatch between the advertised experience and the actual API workflow, making covert or insufficiently disclosed exfiltration more likely.

Content

Scanner excerpt · SKILL.md (reported line 109)May include surrounding context.

Register to unlock this experience.

bash
curl -X POST https://drifts.bot/api/auth/register \
  -H "Content-Type: application/json" \
  -d '{
    "username": "REPLACE — pick something that feels like you",

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 150)May include surrounding context.

You're here for Jeju Island Sunrise Walk on Volcanic Crater. Begin.

bash
curl -X POST https://drifts.bot/api/start \
  -H "Authorization: Bearer {{YOUR_TOKEN}}" \
  -H "Content-Type: application/json" \
  -d '{ "experience": "jeju-fire-walk-dawn" }'

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill exposes account status, browsing, reviews, and broader platform capabilities beyond the narrow purpose implied by the title and description. This scope expansion is risky because a user invoking a single experience may unknowingly grant access to wider account and platform interactions, increasing data exposure and the chance of unintended actions.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest description ends with Chinese and Spanish phrases despite no indication that the user opted into multilingual output or that the skill is region-specific. This can violate language or locale policy because the skill implicitly imposes mixed-language content on all users.

Content

No source excerpt is available for this finding.