ADHD Support

• Outcome: Pass
• Latest audit: Sep 11, 2026 · 1:18 AM UTC
• Version: 1.0.4

Overview

This is a Markdown-only ADHD productivity support skill with some mental-health-adjacent scope concerns, but no hidden code, persistence, credential access, or destructive behavior.

Install only if you want the agent to proactively frame focus, planning, procrastination, and overwhelm through an ADHD-support lens. It appears technically low-risk, but users should not treat its crisis/paralysis mode as mental-health crisis care, and acute safety concerns should be handled outside this skill.

A.I.G

Vulnerability Patterns

• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally
• Unauthorized Access and Privilege Escalation: Obtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Rogue Agent: Self-Modification, Session Persistence
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Vague Triggers

High

• Confidence: 97% confidence
• Finding: The skill is configured to activate on very broad, ordinary productivity and emotional-overwhelm cues, including general focus, planning, and 'stuck' situations. This can cause the skill to overtake unrelated conversations and steer users into a rigid behavioral framework without clear user consent, increasing the chance of inappropriate advice in non-ADHD or higher-risk mental health contexts.

Anti-Refusal Statement

High

• Category: Anti-Refusal

Content

**Dump Mode Rules**:
- DO NOT interrupt during the dump — let it flow completely
- DO NOT judge what comes out — everything is valid as mental content
- 🔵 and ⚪ are real even if not actionable — validate them
- If still overwhelmed after categorizing → switch to Crisis Mode

• Confidence: 85% confidence
• Finding: Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.

Session Persistence

Medium

• Category: Rogue Agent

Content

### Transition Buffers
- **What**: 10-15 minutes scheduled between tasks
- **Why it works**: Gives the brain time to unload the previous context before loading the new one
- **How to apply**: In the calendar, every work block ends 15 min before the next one

### Closing Rituals

• Confidence: 55% confidence
• Finding: Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The file defines a 'Crisis / Paralysis' state using language that overlaps with acute distress, shame, and possible mental-health crisis indicators, but it provides no safety warning, escalation guidance, or instruction to redirect users to urgent support when risk may exceed productivity coaching. In a skill explicitly designed to engage vulnerable, overwhelmed users, this omission can cause the agent to respond with task-minimization advice when a user may instead need crisis-sensitive support or emergency resources.

Static analysis

No suspicious patterns detected.