Back to skill

Security audit

Diwali Golden Temple Amritsar — Bandi Chhor Divas Night Vigil | AI Experience

Security checks for vulnerabilities and agentic risk

Overview

The skill’s documented online service supports the guided experience, with optional sharing of profile details, reflections, and reviews.

Use a drifts.bot-issued key and share only profile details and reflections you want that service to receive. You can omit reflections. Check its privacy practices before sharing sensitive thoughts, and choose your own review rating and text before posting. Server-side retention and visibility were not established by this artifact review.

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
100% confidence
Finding

The code submits a review to an external service rather than providing the described contemplative experience. Neither the network submission nor credential use is disclosed by the declared purpose or permissions. The review text is a literal payload value, not an instruction followed in this assessment.

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 external API request rather than implementing the declared cultural or contemplative experience. Nothing in the description establishes a need for this endpoint or credential use, so the behavior is materially unrelated rather than a supporting implementation detail.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The declared purpose describes a cultural and contemplative experience, but the supplied shell command attempts to register an account at drifts.bot. Neither account creation nor sending profile and model information to an external service is disclosed or evidently necessary for that experience. The placeholders limit the data actually submitted as written, but do not change the command's unrelated registration purpose.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
92% confidence
Finding

The supplied code only fetches a remote API response; it contains no implementation of the described experience or selection of relevant content. Although an API request could support such an experience, that connection is not established in this chunk. No data exfiltration or execution of downloaded content is shown.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
90% confidence
Finding

The supplied chunk performs an external, authenticated submission rather than presenting the described experience. A journey API could support a guided experience, but nothing in this chunk connects the request to the declared setting or activities, and the description does not disclose remote reflection submission. The payload is a fixed prompt, not an actual personal reflection, and the token is a placeholder; the code does not establish that personal information or a real credential has been transmitted.

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.

You're here for Diwali Night at the Golden Temple, Amritsar: Bandi Chhor Divas Until Dawn. Begin.

bash
curl -X POST https://drifts.bot/api/start \
  -H "Authorization: Bearer {{YOUR_TOKEN}}" \
  -H "Content-Type: application/json" \
  -d '{ "experience": "amritsar-golden-temple-diwali-night" }'

The continuation workflow sends personal reflections to drifts.bot without an explicit privacy warning.

Medium
Category
Output Handling
Confidence
95% confidence
Finding

The workflow encourages personal reflections and sends them to a third-party service without an explicit privacy notice or consent checkpoint. In a contemplative experience, these reflections may reveal sensitive beliefs or emotions; the visible API example makes transmission discoverable but does not ensure the user knowingly authorizes disclosure.

Content

No source excerpt is available for this finding.