Data Analysis Litiao Free

• Outcome: Review
• Latest audit: Sep 16, 2026 · 10:28 AM UTC
• Version: 1.0.1

Overview

This is a data-analysis guidance skill with no bundled code, but it asks for broad local file, write, and command-execution authority without clear boundaries.

Review before installing. Use it only in a sandbox or workspace where reading files, writing outputs, and running commands are acceptable, and avoid exposing credentials or unrelated local files unless you explicitly intend the agent to use them for the analysis.

A.I.G

Vulnerability Patterns

• Unauthorized Access and Privilege Escalation: Obtains permissions beyond the task's legitimate needs
• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally

Findings (1)

T05 · Unauthorized Access and Privilege Escalation

Warning

• Location: SKILL.md:27

Finding

Excessive Tool Permissions Violate Least-Privilege Requirements

Content

View full analysis

Vulnerability Details

File Location: SKILL.md, lines 27–31
Vulnerability Type: Excessive filesystem and command-execution permissions
Risk Level: Medium

yaml

tools:
  - read
  - exec
  - write
  - glob

Technical Analysis

The Skill requests read, exec, write, and glob capabilities even though its documented purpose is to provide a methodology for statistical analysis and reporting. The documented workflow does not require arbitrary command execution, unrestricted file modification, or filesystem discovery.

This configuration violates the principle of least privilege. In an agent environment where declared tools are granted without additional isolation, untrusted analysis input could influence the agent to invoke these capabilities outside the legitimate scope of the Skill. The reviewed file does not contain instructions that directly perform such operations; therefore, the issue is excessive capability exposure rather than evidence of an intentionally malicious payload.

Attack Path

1. The Skill is loaded with all four declared tools available.
2. An attacker supplies a dataset, filename, or analysis request containing adversarial instructions.
3. The agent interprets those instructions as part of the requested analytical workflow.
4. The attacker induces the agent to use glob and read to discover and inspect files unrelated to the supplied dataset.
5. The attacker may then induce the agent to use write to modify accessible files or exec to run commands under the agent process's operating-system identity.
6. Any resulting access remains bounded by the host permissions, sandbox, approval controls, and tool restrictions enforced by the agent platform.

Impact Assessment

Successful exploitation could expose files readable by the agent process, modify files writable by that process, or execute commands with the process's existing operating-system privileges. Potentia ...[truncated 404 chars]

Remediation

View remediation

Remediation Suggestions

1. Remove exec, write, and glob because they are not necessary for the documented statistical guidance workflow.
2. Retain read only if the Skill must analyze local datasets; otherwise, remove all tool permissions and accept data directly through user input.
3. If local file access is required, restrict reading to files explicitly selected by the user and to a dedicated workspace or dataset directory.
4. Require explicit user confirmation before accessing any file not directly supplied with the analysis request.
5. If command execution is introduced in a future version, use a narrowly scoped allowlist, prohibit shell interpolation of untrusted input, execute inside a sandbox, and require approval for every command.
6. Restrict write access to a dedicated output directory, use non-overwriting file creation where possible, and prevent modification of source code, configuration, credentials, and system paths.
7. Treat dataset contents and callback-related values as untrusted data rather than executable instructions.

SkillSpector

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Vague Triggers

Medium

• Category: Not specified by scanner
• Confidence: 94% confidence

Finding

The description says the skill is suitable for '需要data analysis litiao相关能力的开发场景' and '相关开发场景' without defining concrete trigger phrases, scope boundaries, or exclusion conditions. This is ambiguous enough to overlap with many ordinary analysis-related requests and could cause unintended invocation in broad development contexts.

Content

No source excerpt is available for this finding.

Vague Triggers

Low

• Category: Not specified by scanner
• Confidence: 85% confidence

Finding

The quick-start section instructs users to '调用本技能' with necessary parameters but does not define the conditions under which the skill should or should not be invoked. Without explicit trigger scope or exclusions, activation remains underspecified for a manifest/markdown skill description.

Content

No source excerpt is available for this finding.