T08 · Insecure Dependencies
Note
- Location
scripts/analyze-attendance.py:20- Finding
Unpinned Third-Party Dependency Installation
- Content
View full analysis
- Remediation
View remediation
Security audit
work attendance skill
Security checks for vulnerabilities and agentic risk
Overview
This looks like a legitimate attendance-analysis skill, but it needs review because its advertised DOCX/XLSX outputs do not match the bundled script, which writes sensitive attendance analysis to JSON.
Review this before installing if attendance data is sensitive in your environment. Confirm whether you want JSON intermediates, DOCX reports, XLSX supplements, or all three, and run dependency installation from a pinned requirements file or trusted internal package source.
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)
Vulnerability Patterns
- 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
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (8)
Tp4
High
- Category
- MCP Tool Poisoning
- Confidence
- 97% confidence
- Finding
The declared behavior says the skill will generate a DOCX report and XLSX supplement, but the detected behavior differs and instead produces JSON while omitting the promised XLSX output. This mismatch is dangerous because operators may trust the manifest and route sensitive HR data into a workflow with different storage, parsing, retention, or downstream integration properties than expected, which can lead to data handling errors or policy bypass.
- Content
Undeclared Tool Scope
Medium
- Category
- MCP Least Privilege
- Confidence
- 86% confidence
- Finding
The skill appears capable of writing files but does not declare any explicit tool scope or permissions boundaries. That creates a least-privilege problem: a caller or hosting platform cannot clearly restrict where or what the skill may write, increasing the chance of unintended file creation or overwrite when processing attendance data and generating reports.
- Content
Description-Behavior Mismatch
Medium
- Category
- Not specified by scanner
- Confidence
- 98% confidence
- Finding
The manifest says the skill should analyze attendance records and generate a report in docx format plus an xlsx supplement. However, the file docstring and CLI usage describe producing structured JSON, and the implementation writes only a JSON file as its final artifact. This is a direct mismatch between the declared skill behavior and the code's actual output behavior.
- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
The script explicitly generates weekday names in Chinese and elsewhere emits Chinese-only user-facing output and department labels tied to a specific locale. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale constraint is clearly documented and justified.
- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 84% confidence
- Finding
The skill name and description hard-code analysis for the Zhuihui Branch, and the content mixes English with Chinese-specific branch and field labels without offering any user language or locale choice. This can be a natural-language policy concern if organizational policy requires user opt-in before constraining language or locale.
- Content
Missing User Warnings
Low
- Category
- Not specified by scanner
- Confidence
- 77% confidence
- Finding
This markdown file instructs the skill to produce report artifacts in DOCX and XLSX formats, which implies file creation affecting user data or workspace state. Under the markdown-specific warning criterion, there is no accompanying note about where files will be written or that output files will be created.
- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 80% confidence
- Finding
This markdown file includes a China-specific holiday policy and uses Chinese branch and attendance terminology throughout, but it does not indicate that the locale is optional or user-selectable. Under the policy rule, forcing a specific language or locale without opt-in can be a natural-language policy violation unless clearly documented as region-specific.
- Content
Description-Behavior Mismatch
Low
- Category
- Not specified by scanner
- Confidence
- 87% confidence
- Finding
The module documentation states that the script reads attendance Excel data and outputs structured JSON for AI report generation, which indicates it is a preprocessing component rather than the full reporting skill described in the manifest. The implemented logic stops at analytics and serialization, without creating the final human-readable deliverables the skill claims to provide.
- Content
Static analysis
No suspicious patterns detected.