Back to skill

Security audit

Phy Proto Break Check

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed protobuf compatibility checker that reads proto files or git history and reports breaking changes without persistence or unrelated data access.

Installers should understand that this checker may read .proto files from the selected directories or git refs and may run local git commands to inspect history. Its main risk is imperfect detection accuracy, so treat findings as a compatibility aid rather than a complete protobuf verifier.

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
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The manifest and detector table explicitly state that SM003 covers "Service name or package changed," implying rename detection for services. In the implementation, SM003 is only emitted for package changes or full file removal, while service name changes are treated as a removed service plus a new service rather than an actual rename detection.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

This markdown skill description shows invocation examples followed by free-form requests like "Compare proto files" and "Check if my .proto changes..." without defining a constrained trigger vocabulary or exclusion conditions. That makes activation scope ambiguous and increases the chance the skill could be invoked by ordinary protobuf-related conversation rather than an intentional command.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
96% confidence
Finding

The inline comment says "Entire message removed — check if any service references it," which describes conditional logic based on service usage. The actual code immediately appends a breaking change for every removed message and performs no reference analysis, so the comment actively misstates what the code does.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.