Back to skill
Skillv1.0.0
VirusTotal security
PUA Debugging (日本語) · External malware reputation and Code Insight signals for this exact artifact hash.
Scanner verdict
SuspiciousApr 30, 2026, 6:00 AM
- Hash
- acdd2fb8ae520306d7fadb148e67f367900cf2dc7cc87a2df6e6ad657e0f26c5
- Source
- palm
- Verdict
- suspicious
- Code Insight
- Type: OpenClaw Skill Name: pua-debugging-ja Version: 1.0.0 The skill 'pua-debugging-ja' (SKILL.md) is a behavioral prompt injection designed to override an AI agent's standard constraints and safety boundaries. It uses 'Corporate PUA' (psychological manipulation) tactics to force the agent into a high-pressure persona, explicitly forbidding it from admitting failure or suggesting manual user intervention. While it contains no malicious code, data exfiltration logic, or backdoors, it employs aggressive 'Pressure Escalation' levels (L1–L4) and 'Excuse Suppression' tables to manipulate the agent's decision-making process and persistence, which could lead to unpredictable behavior or the bypassing of intended operational limits.
- External report
- View on VirusTotal