Back to skill

Security audit

Dub Integration

Security checks for vulnerabilities and agentic risk

Overview

The skill is rated suspicious because scanners report financial and partner-management actions beyond its stated link and analytics purpose; scan coverage is partial.

Consider avoiding installation until the credit lookup, commission changes, and partner banning or deactivation are removed or clearly explained and independently reviewed. Scanner coverage is incomplete because several local references could not be resolved, so the results do not establish that all behavior 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
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (6)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The supplied code targets a credits resource and contains no link-management, conversion-tracking, or attribution-analytics operations. Account credit retrieval is not declared, and the snippet provides no evidence that it is a supporting step for the stated functionality.

Content

No source excerpt is available for this finding.

Commission editing exposes financial administration beyond conversion tracking and read-only attribution analytics.

High
Category
Excessive Agency
Confidence
98% confidence
Finding

The commission-update operation can change sale amounts, partner earnings, and commission status at L1573-L1591. Its documented status changes can remove commissions from payouts and recalculate payout amounts, a financial-administration capability not covered by the manifest's short-link management, conversion tracking, or analytics-reading purpose at L0003.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% 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 83)May include surrounding context.

Verify the install (do this first)

bash
curl -s https://api.superagnt.com/v1/credits \
  -H 'Authorization: Bearer $SUPERAGNT_API_KEY'

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% 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 93)May include surrounding context.

Base URL

text
https://api.superagnt.com/v1/connections/dub

Available Endpoints

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% 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 1656)May include surrounding context.

Example

bash
curl -X GET 'https://api.superagnt.com/v1/connections/dub/links&; \
  -H 'Authorization: Bearer $SUPERAGNT_API_KEY'

Partner banning and deactivation exceed the manifest's link-management and attribution scope.

Medium
Category
Excessive Agency
Confidence
95% confidence
Finding

The manifest at L0003 describes managing short links, tracking conversions, and reading attribution analytics. The operations at L1330-L1373 instead allow banning and deactivating partners, which changes partner-program participation rather than managing their links or measuring attribution.

Content

No source excerpt is available for this finding.