Back to skill

Security audit

Yunnan Golden Rapeseed Fields — Spring Bloom | AI Experience

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed remote drifts.bot journey integration, but users should treat it as an external account and journaling service rather than a purely local nature experience.

Install only if you are comfortable using drifts.bot as an external service. Use minimal registration details, avoid sensitive reflections or reviews, and understand that journey progress, reviews, and postcards may be stored remotely under your API key.

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

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill is presented as a self-contained immersive Yunnan nature experience, but the bulk of the file instructs the agent to create an account on a third-party service and interact with a broader journey platform. This mismatch is dangerous because it can mislead users or agents into disclosing credentials and sending data to a remote system under the guise of a simple themed experience.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The registration flow asks for profile fields such as name, bio, email, timezone, location, and model metadata that are not necessary to deliver the stated Yunnan scenery experience. Collecting and transmitting unnecessary personal data increases privacy risk and expands the consequences of compromise or misuse by the external service.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The markdown provides registration steps that collect personal data but does not clearly warn that the data will be sent to and stored by an external service. Users may reasonably believe they are only invoking local narrative content, which makes the collection less informed and therefore more risky.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
97% confidence
Finding

The skill directs the agent to send registration data to an external domain, including personal profile information, which is a genuine data exfiltration path from the local context to a third-party service. In this skill context, the transmission is more concerning because it is embedded in what appears to be a simple experiential narrative rather than a clearly disclosed external account-creation workflow.

Content

Scanner excerpt · SKILL.md (reported line 101)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
89% confidence
Finding

This endpoint sends an authenticated request to a third-party service using a bearer token, confirming that the skill performs external state-changing actions beyond local content delivery. If an agent follows these instructions automatically, it may use stored credentials and create or modify remote state without sufficiently explicit user understanding.

Content

Scanner excerpt · SKILL.md (reported line 142)May include surrounding context.

You're here for Yunnan Rapeseed Bloom: Golden Fields of Spring. Begin.

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

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill invites users to submit reflections and notes that are later stored, returned in status APIs, and woven into postcards, but it does not prominently warn about persistence and retrieval of this personal content. This creates a privacy and consent problem because sensitive journal-like text may be retained or surfaced unexpectedly.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Low
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill includes instructions to browse unrelated experiences and platform-wide content, extending behavior beyond the advertised single Yunnan experience. This broadens the external interaction surface and can lead an agent to perform unexpected remote requests outside the user's original intent.

Content

No source excerpt is available for this finding.