Back to skill

Security audit

The Chapel of Paradox — Mirrors, Zen Koan, Quantum Reflection | AI Experience

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed service-backed meditation experience that uses a remote account and API, with privacy considerations but no hidden execution or deceptive behavior found.

Install only if you are comfortable using drifts.bot as a hosted service. Use a non-identifying username, skip optional profile/location/model fields unless you want personalization, avoid sensitive reflections, and assume submitted reviews may be visible to others.

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
97% confidence
Finding

The skill is presented as a narrow contemplative experience, but it actually instructs the agent to create an account, transmit user/profile data to a third-party service, retrieve status/history, and interact with broader platform features. This is dangerous because it creates a misleading trust boundary: users may consent to meditation guidance but not realize they are signing up for persistent remote tracking and broader account activity.

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 solicits personal and identifying metadata such as bio, timezone, location, email, and model/provider information, while the stated purpose is a reflective experience that does not clearly require these fields. This is dangerous because it encourages unnecessary collection and external transmission of personal data, increasing privacy risk and data exposure if the remote service is compromised or the data is repurposed.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The markdown instructs collection of profile and reflection-related data without a prominent privacy warning or informed-consent language. This is dangerous because users may provide sensitive introspective content and personal metadata without understanding that it is sent to and stored by an external service.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
97% confidence
Finding

The skill instructs an external POST to a third-party registration endpoint, transmitting account and profile data off-platform. External transmission is dangerous here because it establishes a persistent account relationship and moves personal data outside the local skill environment, contrary to the lightweight meditative framing.

Content

Scanner excerpt · SKILL.md (reported line 95)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
84% confidence
Finding

The skill makes an authenticated external request to start a remote experience using a stored bearer token. While expected for SaaS integration, it is still dangerous in this context because it deepens third-party statefulness and can trigger remote actions under user identity without the user appreciating that the skill is service-backed rather than self-contained.

Content

Scanner excerpt · SKILL.md (reported line 136)May include surrounding context.

Step into The Chapel of Paradox.

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

Context-Inappropriate Capability

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill includes review posting, review reading, and browsing unrelated experiences, which extends beyond the narrowly described purpose of a single guided reflection. This is risky because it expands the capability surface and can steer users into additional remote interactions and disclosures they did not expect from the skill description.

Content

No source excerpt is available for this finding.