Back to skill

Security audit

Command Compass Daily File Index

Security checks for vulnerabilities and agentic risk

Overview

The core skill is local and privacy-conscious, but the package also contains instructions to change website APIs, account storage, and a downloadable skill ZIP.

Review this package before installing. It is suitable as a local card-generation skill only if you ignore or remove the website-agent instructions; do not let an agent apply the website/API/storage or downloadable ZIP replacement steps unless you intentionally want website maintenance changes and have a separate approval/release process.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (10)

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The instructions clearly direct the agent to modify a remote website codebase, API behavior, and downloadable artifacts, which exceeds the stated scope of a local-first resource indexing skill. This kind of scope expansion is dangerous because it can cause the skill to operate on unrelated systems and deploy changes to public-facing infrastructure without an appropriate trust boundary or user approval.

Content

No source excerpt is available for this finding.

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
80% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · WEBSITE_AGENT_INSTRUCTIONS.md (reported line 140)May include surrounding context.

md
- `GET /api/favorites`
- `POST /api/favorites`
- `DELETE /api/favorites`

要求:

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The entire report is written in Chinese and provides no indication that language selection is optional or constrained by a documented region-specific requirement. Per the policy, forcing a specific language without user opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill mandates Windows client category names that are exclusively Chinese (for example 技能, 文本, 图片) and does not indicate that users may choose another language or locale. This is a natural-language locale constraint that can violate policy when imposed without explicit opt-in or justification.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

This markdown file contains all operational instructions in Chinese and does not indicate that language selection is optional or user-configurable. Under the policy rule for language/locale constraints, forcing a specific language without opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

These instructions add authenticated remote account features, favorites management, hound APIs, and server-side storage design that are not justified by the skill's local-first indexing purpose. Even if framed as product work, this broadens the agent's operational authority into handling authentication and user data flows, increasing the risk of unauthorized changes, privacy mistakes, or cross-account data exposure.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

Replacing the website's downloadable skill package is a deployment and supply-chain action unrelated to local resource indexing. This is dangerous because it gives the skill authority to alter distributable artifacts consumed by end users, creating a path for unintended or malicious package substitution.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

This manifest uses Chinese-only natural-language fields such as title, summary, tags, category, and instruction, and there is no indication that the skill is region-specific or that users can opt into another language. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

This card's user-facing text is entirely in Chinese and the manifest does not state that the skill is limited to a Chinese locale or audience. Because no opt-in or alternative language is provided, it violates the language/locale policy criterion.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The title, summary, tags, category, and instruction for this card are presented only in Chinese, with no evidence that users can choose language or that the card is intentionally constrained to a Chinese-speaking region. That fits the language/locale policy violation definition.

Content

No source excerpt is available for this finding.