Back to skill

Security audit

Li Git1.1 Checker

Security checks for vulnerabilities and agentic risk

Overview

This is a straightforward standards-document checker with some scope and privacy caveats, but no hidden persistence, exfiltration, or destructive behavior was found.

Install this if you need GB/T 1.1-2020 standards-structure review. Use explicit prompts and avoid submitting confidential drafts unless you are comfortable with the document being analyzed visually; treat the international-standards sections as approximate guidance rather than authoritative compliance checking.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (9)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The description is written as a definitive Chinese-only capability for the skill and does not mention any user-selectable language or locale option. Under the policy, forcing a specific language without opt-in is a natural-language policy violation unless the locale constraint is explicitly documented and justified as region-specific.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The trigger phrase and invocation description are Chinese-only and do not indicate that the skill is limited to Chinese-speaking users or PRC standardization workflows. Because the file does not clearly justify the language restriction or offer a choice, this reads as a mandatory language/locale constraint.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill includes operational instructions to scan a repository for hardcoded paths, usernames, and secrets before upload, which is outside the stated purpose of GB/T 1.1 document-structure checking. This broadens the skill into repo-inspection and upload-preparation behavior, increasing the chance an agent will access unrelated files or expose sensitive metadata during normal use.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest says this is a '基于 GB/T 1.1-2020 标准文档结构检查工具', which implies a checker scoped to the GB/T 1.1-2020 standard. However, the file presents the skill as usable across ANSI/ASTM, CEN/CENELEC, JIS, DIN, BS, AFNOR, and ISO/IEC systems, broadening its behavior beyond the stated GB/T-specific scope.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The guide states that Phase 2 uses visual reading of document content via 'vision_analyze' but provides no user-facing notice about data handling, retention, or sensitivity considerations. Because standards drafts may contain confidential or prepublication material, silent image-based analysis can expose users to unanticipated privacy, confidentiality, or policy violations.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The manifest frames the skill as a GB/T 1.1-2020 document structure checker, but these lines state it can '直接用于 ISO 标准结构检查' and even equate GB/T compliance checking with ISO/IEC Directives compliance. That is a substantive scope expansion in claimed behavior, not just an implementation detail of checking GB/T documents.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The phrase '用户:请检查这个文档' followed by automatic skill loading creates an overly broad trigger that can cause the skill to activate on generic document-review requests without informed user intent. In practice this raises the chance of unintended processing of sensitive files and misrouting of user tasks to a capability that performs document inspection beyond what the user specifically requested.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

Documenting automatic skill loading for uploaded documents without a warning about unintended processing creates a real privacy and user-consent risk. A user may provide a document for general assistance, yet the system could automatically route it into structured analysis and visual inspection workflows they did not knowingly authorize.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

This markdown file is natural-language instructional content, and its title, headings, tables, and code-facing output strings are all presented in Chinese. Under the policy rule, forcing a specific language without user opt-in can be a locale/language policy violation unless the constraint is explicitly documented and justified.

Content

No source excerpt is available for this finding.