Security audit

灵感素材整理器

Security checks for vulnerabilities and agentic risk

Overview

This skill locally saves and organizes user-provided inspiration notes, with no evidence of network access, credential access, hidden execution, or destructive behavior.

Install only if you want a Chinese-language local inspiration-note organizer. Be aware that captured text is stored in a local JSON file, and do not set INSPIRATION_DB_PATH or INSPIRATION_TOPIC_RULES_PATH to important unrelated files.

A.I.G

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

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
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 (9)

Undeclared Tool Scope

Medium

Category: MCP Least Privilege
Confidence: 94% confidence
Finding: The skill instructs use of a Python script and documents environment-variable-controlled paths plus local JSON read/write behavior, but it declares no explicit tool scope or permissions. That creates a capability/authorization mismatch: an agent may invoke file and environment access without transparent limitation, increasing the chance of unintended reads/writes or path abuse if the implementation trusts overridden paths.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 88% confidence
Finding: The activation criteria are very broad and match common note-taking, search, and organization requests without clear exclusion boundaries. This can cause the skill to trigger in more contexts than intended, leading an agent to collect, classify, or persist user content unnecessarily, which expands privacy exposure and increases the chance of unintended file operations.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: This JSON manifest defines topic categories using many generic single-word triggers such as “用户”, “功能”, “学习”, “沟通”, and “品牌” without any scope limits, exclusion rules, or negative examples. Because these terms are common in ordinary conversation, they create ambiguous activation boundaries and may cause the skill to trigger in contexts unrelated to the intended domain classification.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 84% confidence
Finding: All category names and matching terms are hard-coded in Chinese, and the file provides no indication that users may choose another language or that the skill is intentionally limited to a Chinese-language environment. This can violate language or locale policy when the skill is used in broader multilingual contexts without explicit opt-in.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 89% confidence
Finding: This markdown file presents all user-facing instructions in Chinese and does not indicate that users may choose another language or that the skill is intentionally limited to a Chinese-speaking context. Under the language/locale policy, forcing a specific language without user opt-in is a natural-language policy concern.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The parser description, subcommand help, argument help, and runtime output strings are all fixed in Chinese with no option to select another language. This creates a natural-language locale constraint without user opt-in or documented justification.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 91% confidence
Finding: This JSON content contains user-facing title, content, topic, tags, and group fields entirely in Chinese, but provides no natural-language indication that the skill is Chinese-only, region-specific, or offers language choice. Under the language/locale policy rule, forcing a specific language without opt-in can be a policy concern.
Content: No source excerpt is available for this finding.

Missing User Warnings

Low

Category: Not specified by scanner
Confidence: 89% confidence
Finding: The code persists captured or classified card data by overwriting the JSON database file, but this function itself provides no explicit warning that local data will be modified. Although callers print success messages after completion, there is no prior disclosure or confirmation around the write operation.
Content: No source excerpt is available for this finding.

Missing User Warnings

Low

Category: Not specified by scanner
Confidence: 85% confidence
Finding: When the topic-rules file does not exist, the script silently creates directories and writes a default configuration file. This is a filesystem-modifying action that lacks a visible warning, prompt, or explanatory comment to inform the user that configuration files will be generated.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.