Back to skill

Security audit

Pans Pricing Engine

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward GPU pricing calculator, but users should validate inputs before relying on generated quotes.

Install only if a Chinese-language GPU pricing calculator fits your workflow. Treat its output as advisory unless positive count and duration validation is added, especially before feeding JSON quotes into business or billing systems.

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/pricing.py:90
Finding

Missing Validation Allows Negative Pricing Quotes

Content
View full analysis
PricingResult: """计算最终定价""" if gpu not in BASE_PRICES: raise ValueError(f"不支持的GPU型号: {gpu},支持的型号: {list(BASE_PRICES.keys())}") base_price_per_hr = BASE_PRICES[gpu] base_monthly = base_price_per_hr * count * 730 # ~730小时/月 base_yearly = base_monthly * 12 q_discount = get_quantity_discount(count) d_discount = get_duration_discount(duration_months) p_discount = get_prepay_discount(prepay_cycle) # 累乘折扣 (1 - d1) * (1 - d2) * (1 - d3) combined = 1.0 - (1.0 - q_discount) * (1.0 - d_discount) * (1.0 - p_discount) discounted_monthly = base_monthly * (1.0 - combined) discounted_yearly = discounted_monthly * 12 return PricingResult( gpu=gpu, base_price_per_hr=base_price_per_hr, gpu_count=count, duration_months=duration_months, prepay_cycle=prepay_cycle, base_monthly=base_monthly, base_yearly=base_yearly, quantity_discount=q_discount, duration_discount=d_discount, prepay_discount=p_discount, combined_discount=combined, discounted_monthly=discounted_monthly, discounted_yearly=discounted_yearly, monthly_savings=base_monthly - discounted_monthly, yearly_savings=base_yearly - discounted_yearly, price_per_gpu_hr=base_price_per_hr * (1.0 - combined), ) ``` The same unvalidated values reach this function from the command-line path: ```python result = calculate_pricing(args.gpu, args.count, args.duration, args.prepay) ``` The comparison path also parses signed integers without validating their permitted ranges: ```python try: count = int(parts[matched_n]) duration = int(parts[mat ...[truncated 2634 chars]
Remediation
View remediation
Vulnerability Patterns
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (5)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The phrase "quote generator" is ambiguous and generic, so it may activate on many unrelated requests involving quotations, proposals, or document generation. This increases the risk of accidental invocation and misrouting, especially in multi-skill environments where generic English triggers can collide with broader sales or document workflows.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The phrase "quote generator" is ambiguous and generic, so it may activate on many unrelated requests involving quotations, proposals, or document generation. This increases the risk of accidental invocation and misrouting, especially in multi-skill environments where generic English triggers can collide with broader sales or document workflows.

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
82% confidence
Finding

The trigger list lacks clear non-trigger conditions and activation boundaries, making it unclear when the skill should or should not be selected. In practice this can lead to opportunistic or inconsistent routing, increasing the chance that ordinary pricing discussions are mistaken for requests to run this specialized contract-pricing engine.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The module docstring presents the skill description only in Chinese, and the rest of the CLI help/output is likewise Chinese-only. This can violate a language-choice policy when the skill forces a specific language without giving users an opt-in or alternative locale.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
97% confidence
Finding

User-visible strings for pricing listings, comparison output, argument help, examples, and error messages are all fixed in Chinese. The file does not indicate that the tool is region-specific or provide any mechanism for users to choose another language.

Content

No source excerpt is available for this finding.