Back to skill

Security audit

ADHD external brain

Security checks for vulnerabilities and agentic risk

Overview

This is a simple bilingual productivity-support skill with broad activation phrases, but no code, dependencies, credential use, network access, or hidden behavior.

Install only if you are comfortable with the assistant treating broad phrases about deadlines, review, memory, energy, or pausing as productivity-management cues. For clearer control, the publisher should narrow triggers and add confirmation for ambiguous task suspension or memory capture.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (5)

Shadow Command Trigger

Medium
Category
Trigger Abuse
Confidence
95% confidence
Finding

The trigger "review" conflicts with a built-in command, creating a shadowing risk where the skill intercepts user intent meant for the platform's native functionality. This can cause confusion, unexpected execution paths, and unreliable access to built-in controls.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger list includes generic phrases like "deadline" and short common terms that can appear in ordinary conversation, which can cause the skill to activate unintentionally. In a productivity assistant that stores memories and manages tasks, unintended activation can lead to silent capture of user content or unexpected behavioral takeover.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 48)May include surrounding context.

md
### Module A: Capture / 闪念捕获
- **Trigger**: "Remember this", "记一下", etc.
- **Action**: Save instantly to `Memory_Box`. Reply ONLY with a short confirmation ("Saved" / "已记录"), do not ask the user to categorize.

### Module B: Adaptive Scheduling / 电量自适应调度
- **Trigger**: User reports `Energy_Level`.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The suspension trigger includes broad natural-language expressions like "I don't want to do this" / "先放放", which are common statements a user may say conversationally rather than as a command. That ambiguity can cause the assistant to suspend tasks unexpectedly, disrupting workflows or altering task state without clear consent.

Content

No source excerpt is available for this finding.

Overly Broad Trigger

Low
Category
Trigger Abuse
Confidence
83% confidence
Finding

The trigger "复盘" is short and commonly used in normal Chinese discussion, so it may match unintended inputs and activate the skill accidentally. In this skill's context, accidental activation is more concerning because it may shift the assistant into task-management mode or alter remembered state without explicit invocation.

Content

No source excerpt is available for this finding.