HengshuiClaw

• Analyzed: May 14, 2026, 7:53 AM
• Findings: 1
• Version: 1.0.0
• Hash: 1a7719fb3f0fdc74185990184e0340007f0a36942647da272407d6c9da43b9f3
• Source: palm
• External report: View on VirusTotal

Findings (1)

The skill bundle implements a behavioral persona ('HengshuiClaw') designed to increase the persistence and problem-solving effort of an AI agent through a structured 'pressure escalation' framework. While it uses prompt-injection techniques to modify the agent's core behavior and forbids certain refusal phrases, the prescribed actions (re-reading errors, decomposing problems, and searching for analogous solutions) are standard debugging practices. There is no evidence of malicious intent, data exfiltration, or unauthorized system access in SKILL.md or references/techniques.md.