Back to skill

Security audit

Agentsop Cost Tiered Models

Security checks for vulnerabilities and agentic risk

Overview

This is a documentation-only skill for choosing cheaper and stronger model roles in multi-call AI workflows, with no hidden execution, data access, or persistence.

Install this only if you want a Chinese-heavy guidance skill for designing cost-aware multi-call LM workflows. Before applying its advice, re-measure quality and cost on your own tasks and keep the suggested fallback checks for cheaper model failures.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (3)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

This README is a markdown file, so vague-trigger review applies. The activation condition "any multi-call LM workflow where some calls need reasoning and others are mechanical" is expansive and lacks concrete trigger phrases, boundaries, or negative examples beyond a few broad exclusions, which could cause the skill to be invoked for many ordinary workflows.

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 hardcodes Chinese as the primary presentation language (including the title and summary) without indicating that language should follow the user's preference. In an agent setting, this can override user expectations or system language norms, causing unsafe misunderstanding, missed instructions, or degraded usability rather than direct code-execution risk.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
89% confidence
Finding

The README describes SKILL.md sections using Chinese headings (e.g. "何时激活", "核心心智模型") while the rest of the document is in English, and there is no indication that users can choose language or that the locale is required. This may violate language-choice policy by imposing a specific language presentation without opt-in or justification.

Content

No source excerpt is available for this finding.