Back to skill

Security audit

Typhoon Chichijima Ogasawara — Riding Out a Typhoon | AI Experience

Security checks for vulnerabilities and agentic risk

Overview

This skill is a story experience, but it also asks users to create and use a drifts.bot account that sends profile, journey, reflection, and review data to an external service without enough privacy or retention detail.

Install only if you are comfortable creating a drifts.bot account and sending the service profile details, location or timezone if provided, reflections, journey progress, and possibly public review text. Use minimal optional profile data, protect the API token, and look for the service's privacy and deletion terms before using it for personal content.

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 (6)

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The manifest markets the skill as a self-contained immersive experience, but the body reveals collection of profile data, remote account creation, progress tracking, and review posting to an external service. This mismatch can mislead users and agents into sharing data or invoking networked side effects they would not reasonably expect from the description.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill requests an API token and instructs the user to create an authenticated account for what is presented as a narrative/immersive experience. That expands the trust boundary unnecessarily, enables remote state changes, and creates risk of credential misuse, tracking, and unexpected external interactions unrelated to the core descriptive function.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The registration flow solicits personal data such as bio, email, timezone, and location without any visible privacy notice, retention explanation, or minimization guidance. Users may disclose sensitive or identifying information to a third-party service without understanding how it will be stored, used, or shared.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
92% confidence
Finding

This endpoint transmits user-supplied registration data to an external domain, including potentially identifying profile attributes. In the context of a skill framed as an immersive experience, this is a meaningful external data exfiltration/vector for account creation and should be treated as a side effect requiring strong disclosure and consent.

Content

Scanner excerpt · SKILL.md (reported line 106)May include surrounding context.

Create an account to begin traveling.

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
88% confidence
Finding

This call sends an authenticated request to an external service to start a remote journey, creating or mutating server-side state. Even though the action is expected by the API design, it is still a security-relevant side effect because it uses a bearer token and performs networked state changes beyond simple local content rendering.

Content

Scanner excerpt · SKILL.md (reported line 147)May include surrounding context.

This is where Riding Out a Typhoon on Chichijima, Ogasawara Islands begins.

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

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
92% confidence
Finding

The top-level description mixes English, Chinese, and Spanish text without offering the user a language or locale choice. This can violate language/locale policy expectations because the skill presents multilingual content by default rather than documenting or requesting user preference.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.