Back to skill

Security audit

subscription-churn-analysis

Security checks for vulnerabilities and agentic risk

Overview

The skill uses limited local access, but its broken scoring logic makes the generated churn rankings unreliable.

Do not rely on the generated scores for customer-retention decisions until input mappings, missing-data handling, grouping, and renewal ordering are corrected and validated. The optional Windows app is unnecessary for the Python workflow and requires a separate trust decision.

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
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The reported empty-column lookups and omitted risk criteria would compromise report integrity, potentially presenting unusable scores as meaningful and causing teams to overlook at-risk revenue. This contradicts the documented scoring behavior and missing-data guardrail, but the referenced script was not supplied, so the implementation defect cannot be independently verified here.

Content

No source excerpt is available for this finding.

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: file_write.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

The churn scorer does not read any of the declared churn signals, so ordinary account CSVs receive identical zero scores.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

Every declared criterion lacks a column mapping, so the scorer looks up an empty-string CSV header and ordinarily substitutes zero for every churn signal. It then presents identical zero scores as a ranked report, silently undermining the integrity of the analysis and potentially causing customer-success teams to overlook accounts at genuine risk. This is especially consequential for a skill whose core purpose is churn prioritization, although the demonstrated impact is decision integrity rather than code execution or data theft.

Content

No source excerpt is available for this finding.