Back to skill

Security audit

Pans Churn Predictor

Security checks for vulnerabilities and agentic risk

Overview

This skill appears purpose-built for customer churn analysis, but its initialization command can overwrite local customer, usage, ticket, and contract data without confirmation.

Review before installing if this would be used with real customer or contract records. Avoid running --init on an existing installation unless you have backups or are comfortable replacing the skill's local datasets with bundled sample data.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (1)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/churn.py:286
Finding

Unprotected initialization overwrites existing customer datasets

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (8)

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
79% confidence
Finding

The skill advertises executable commands that read and likely write local files under ~/.qclaw/skills/pans-churn-predictor/data/, but the manifest declares no explicit tool scope such as permissions or allowed-tools. This creates an authorization ambiguity where a host agent may invoke file-capable behavior without clear least-privilege boundaries, increasing the chance of unintended data access or modification involving customer usage, tickets, and contract data.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger list includes broad business phrases such as '客户成功' and 'churn prediction', which may cause the skill to activate in loosely related conversations rather than only when the user clearly intends to run this customer-risk workflow. Over-broad activation can expose sensitive customer-health data or lead the agent to perform unintended analysis in the wrong context.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill's natural-language description and user-facing interface strings are in Chinese throughout the file, which effectively forces a specific language for interaction. There is no indication that the user can choose another language or that the locale restriction is explicitly documented as intentional and justified.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The manifest and module/docs describe a customer churn warning and analysis tool, which implies read-oriented analytics over existing business data. However, the code includes an --init path that creates and writes multiple local datasets, turning the skill into a data-seeding tool in addition to analysis.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

The --init path writes multiple local data files without checking for existing content or warning that files may be created or overwritten. In this skill's context, the risk is limited to accidental local data loss or contamination of analyst/customer records, not system compromise, but it is still an unsafe state-changing behavior.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

This JSON file encodes all client names, industries, and contacts exclusively in Chinese, which constitutes a language/locale choice baked into the skill data. There is no indication in the file that users can choose another language or that the locale restriction is explicitly justified.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
80% confidence
Finding

This JSON file uses Chinese-only organization names as the primary visible natural-language content, with no indication that the skill or dataset is intentionally region-specific or that users can choose another language/locale. Under the stated policy, a fixed language/locale can be a violation when it is imposed without opt-in or documented justification.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

The JSON values for ticket issues and organization names are entirely in Chinese, with no indication that language selection is configurable or optional. Under the policy criteria, hard-coding a single language/locale without opt-in can constitute a natural-language policy violation.

Content

No source excerpt is available for this finding.