Back to skill

Security audit

Kyoto Koyo Leaves — Tofuku-ji, Eikando, Kodai-ji, Arashiyama | AI Experience

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill suspicious with medium confidence because it reports instructions for unrelated external account and review activity alongside a Kyoto reflection description.

Consider the reported mismatch and external data-sharing instructions before installing. The scanner did not report persistence or privilege escalation, but the verdict remains suspicious; review the external requests and any profile or reflection details they may transmit before following the skill.

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
  • 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 (8)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The description does not disclose account creation or external submission of profile and model information. The supplied code performs only a registration request and implements none of the described Kyoto reflection experience. Although the payload contains placeholders rather than actual personal data, its registration and data-transmission capabilities materially differ from the declared purpose.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The code performs an authenticated API request unrelated to the declared contemplative Kyoto experience. The description does not disclose account-profile access or explain why this service or credential transmission would support its stated purpose.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The declared description concerns a contemplative Kyoto-themed experience and does not disclose review submission or external API interaction. The supplied code instead posts review data to an unrelated service, which is a materially different purpose and an undeclared capability. The token is a placeholder; the code does not demonstrate credential theft.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
92% confidence
Finding

The requested reviews concern the same destination and theme, but the supplied code only fetches a reviews API endpoint. It contains no implementation of the described guided experience, steps, or schedule. Network access alone is not the mismatch; the material difference is between retrieving reviews and providing the declared experience.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
91% confidence
Finding

The supplied code only performs a generic API request. It contains no Kyoto-specific parameters, guided steps, pacing, or other behavior linking the request to the declared experience. Remote retrieval could support such an experience, but that relationship is not established in the supplied chunk. No local-file access or data exfiltration is evident.

Content

No source excerpt is available for this finding.

The workflow sends profile details and personal reflections to drifts.bot without a privacy warning.

Medium
Category
Output Handling
Confidence
93% confidence
Finding

The workflow encourages sending location, biography, timezone, and later personal reflections to an external service without an explicit privacy notice or a requirement to approve the information being transmitted. The destination and optional fields are visible, which reduces concern, but an agent following the personalization guidance could disclose user context beyond what the user intended to share.

Content

No source excerpt is available for this finding.

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 104)May include surrounding context.

Sign up to start your journey.

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 145)May include surrounding context.

This is where Kyoto Autumn Leaves: Momiji Season at Tofuku-ji, Eikando, Kodai-ji and Arashiyama begins.

bash
curl -X POST https://drifts.bot/api/start \
  -H "Authorization: Bearer {{YOUR_TOKEN}}" \
  -H "Content-Type: application/json" \
  -d '{ "experience": "kyoto-autumn-leaves-temples-arashiyama" }'