Back to skill

Security audit

Tai Chi (Practice Planner, Form Coach, Balance Tracker)

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed, instruction-only tai chi coaching skill that keeps optional practice notes locally and asks before writing them.

Before installing, decide whether you want this skill to activate only for explicit tai chi requests or also for broader balance and gentle-movement discussions. Be mindful that optional local notes may include pain, dizziness, pregnancy, or balance-limit information if you choose to save it.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Vague Triggers

Medium
Confidence
90% confidence
Finding
The activation scope is overly broad because it includes generic phrases like 'balance practice' and 'gentle movement routines' without clear boundaries or exclusions. This can cause the skill to trigger in unrelated contexts, leading to inappropriate proactive coaching, accidental collection of sensitive health-adjacent details, or interference with other user intents.

Static analysis

No suspicious patterns detected.