clausewitz-perspective

Pass

Audited by Gen Agent Trust Hub on Aug 4, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill provides detailed instructions for adopting a specific persona (Carl von Clausewitz) and using a specific theoretical framework. These instructions are focused on cognitive modeling and do not attempt to bypass safety filters or override agent constraints.
  • [DATA_EXFILTRATION]: No network operations (curl, wget, fetch) or attempts to access sensitive local files (~/.ssh, ~/.aws, .env) were detected.
  • [REMOTE_CODE_EXECUTION]: The skill does not include any scripts or commands that download or execute code from remote servers. All referenced files are local to the skill package.
  • [COMMAND_EXECUTION]: No shell command execution patterns or use of dynamic context injection (the !command syntax) were found.
  • [OBFUSCATION]: The content is provided in plain text (using Chinese, English, and German terms). No hidden characters, encoded strings, or suspicious obfuscation techniques were detected.
  • [INDIRECT_PROMPT_INJECTION]: While the skill processes user input, it lacks any high-privilege capabilities (file system writes, network access) that could be exploited via indirect injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 4, 2026, 01:27 AM
Security Audit — agent-trust-hub — clausewitz-perspective