Back to skill

Security audit

meta-industry-chain-analysis

Security checks for vulnerabilities and agentic risk

Overview

The skill is a research workflow helper, but its bundled learner persists user preferences, notes, and cross-skill usage state without clear consent, path limits, or retention controls.

Review this before installing if you do not want a skill to keep local learning state. Avoid putting sensitive business details into learner notes or preferences, and inspect or clear learned_patterns.json periodically. A safer version would make persistence opt-in, restrict writes to one dedicated state path, and document deletion and retention behavior.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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)

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

This is a significant description-behavior mismatch: the skill claims to be an industry-chain research meta-skill, but analysis indicates it also performs local persistence via learned_patterns.json and stores user preferences or operation records. Hidden stateful behavior is dangerous because it can silently collect sensitive user/workflow data, manipulate future outputs, and bypass user expectations about what a research skill should do.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
84% confidence
Finding

The skill appears to require file-write capability but does not declare any tool scope or permissions boundary. That creates an authorization gap: consumers may load a seemingly documentation-only research skill while hidden persistence behavior can write local state, reducing transparency and increasing the chance of unauthorized data retention or tampering.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

This module is explicitly designed as a generic cross-skill telemetry and self-evolution component, despite the skill being described as industry-chain analysis. That mismatch matters because it enables persistent collection of usage history, errors, notes, and preferences across arbitrary skills, expanding data handling and behavior beyond user-expected scope and increasing privacy and governance risk.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

This code file contains natural-language strings that define usage and behavior only in Chinese, including the module docstring and command descriptions. Under the policy, forcing a specific language without opt-in is a natural-language policy violation because users are not given any language or locale choice.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The default state includes persistent storage for user preferences, recent operations, error notes, and optimization history without any visible consent, minimization, or sensitivity controls. Free-form notes can easily capture sensitive business context or personal data, making the persistence of this information risky if the skill is shared, inspected, or reused.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The save path is derived from a user-supplied skill directory and the module writes learned_patterns.json into that location with no validation that it belongs to the current skill boundary. This creates stateful filesystem modification in arbitrary directories, which is dangerous in a multi-skill environment because it can alter other skills' state and persist data where the user may not expect it.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The argparse description and printed status messages are all Chinese-language strings, so users interacting with the CLI receive only one locale. Because no alternative language or opt-in mechanism is offered, this conflicts with the language/locale policy described in SQP-3.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The entire skill document is written in Chinese and presents the report format and labels exclusively in that language, with no indication that the user can choose another language or that the skill is intentionally limited to a Chinese-language context. This can violate a language/locale policy when a skill implicitly forces a specific language without opt-in.

Content

No source excerpt is available for this finding.