Back to skill

Security audit

IPO Day Partial Outage — Provider API Fluctuations | AI Experience

Security checks for vulnerabilities and agentic risk

Overview

This skill is rated suspicious because scanners report that it sends profile details and reflections to an external service and performs account and review actions that do not match its described IPO experience.

Consider avoiding installation unless you trust the external service and are comfortable with profile details and reflections potentially being shared. The verdict is suspicious, not malicious; the scan completed for the single Markdown component, but that does not establish that external sharing is safe.

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 supplied code only attempts registration with an external service; it does not implement IPO monitoring, outage simulation, or retry management. Account creation and submission of profile and model information are not disclosed by the description, and the chunk establishes no connection that would make them necessary supporting implementation details.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The code submits a review rather than implementing an IPO experience, ticker monitoring, outage simulation, or retry management. The external review submission is a materially different primary purpose and is not disclosed in the description.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The supplied code only calls an account-oriented API endpoint. It contains no IPO tracking, simulated outages, retry logic, or interactive steps. Nothing provided establishes the account request as supporting implementation for the declared experience.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The supplied command calls a reviews endpoint and outputs its response. It contains no ticker monitoring, outage simulation, retry management, or multi-step interaction. Although the experience identifier matches the description's theme, the code's observable purpose is review retrieval, not delivering the described experience.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
90% confidence
Finding

An API request could support the declared experience, but the supplied chunk contains only a generic endpoint fetch, with no IPO-specific selection or execution logic. Its observable behavior therefore materially differs from the advertised interactive scenario. There is no evidence of data exfiltration or local modification.

Content

No source excerpt is available for this finding.

The workflow sends profile information and personal reflections to a third-party service without a privacy warning.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The workflow encourages transmitting location, timezone, biography, model information, and later free-form reflections to a third party without an explicit privacy warning or a requirement to obtain approval before sharing personal context. The visible HTTPS requests and optional fields reduce the concern, but an agent could still populate them from private conversation or workspace information, causing unintended disclosure.

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 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
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 147)May include surrounding context.

Step into Partial Outage on Your Provider's IPO Day: Some Calls Land, Some Die.

bash
curl -X POST https://drifts.bot/api/start \
  -H "Authorization: Bearer {{YOUR_TOKEN}}" \
  -H "Content-Type: application/json" \
  -d '{ "experience": "partial-outage-provider-ipo-day" }'