Back to skill

Security audit

导购结构分析

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly does the advertised retail clerk performance analysis, but it loads an unbundled API client from a hard-coded developer machine path before fetching employee sales data.

Install only if you trust the publisher and can verify or replace the missing api_client dependency. Before using it with real store data, confirm what service get_copilot_data contacts, what credentials it uses, and that the user is authorized to process named clerk performance metrics.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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)

T08 · Insecure Dependencies

Warning
Location
analyze.py:7
Finding

Untrusted External Module Loading Through a Hard-Coded Import Path

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (5)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The trigger section lists natural-language intents such as "导购团队结构如何" and "哪些导购对业绩影响最大" without defining exclusions, required retail context, or negative examples. These phrases could overlap with general analytical discussion and may cause unintended invocation outside the intended store-clerk performance analysis scenario.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The module-level natural-language description and user-facing log/output strings are entirely in Chinese, which effectively forces a specific language/locale. The file does not offer any language selection, fallback, or opt-in mechanism.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The code prepends a hard-coded local filesystem path to sys.path, causing imports to be resolved from an uncontrolled developer-specific workspace location. This creates a supply-chain and environment-manipulation risk: if that path exists and contains a modified api_client module, the skill may execute unintended code or access data through an unreviewed dependency.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill fetches store and clerk performance data over the network without any normal user-facing disclosure, consent, or transparency in non-verbose execution. Because this data is individually attributable employee performance information, silent retrieval increases privacy and data-governance risk and can lead to unauthorized or unexpected processing of sensitive business metrics.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

The manifest description and trigger instructions force a specific language presentation, and the document does not state that the skill is region-specific or provide any user opt-in for language/locale. Under the policy, locale constraints should be justified or optional rather than implicit.

Content

No source excerpt is available for this finding.