karl-marx
Pass
Audited by Gen Agent Trust Hub on Apr 30, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill focuses on defining a specific analytical persona and does not contain any malicious code, hidden commands, or persistence mechanisms. The content is derived from historical and philosophical works and is intended for roleplay and analysis purposes.
- [PROMPT_INJECTION]: The instructions establish a strict roleplay persona with a defined workflow for different user 'camps'. These instructions prioritize structural critique over personal attacks and explicitly forbid the AI from insulting or blaming the user, which aligns with standard safety guidelines. It includes a clear 'exit' command for the user to deactivate the persona.
- [EXTERNAL_DOWNLOADS]: The skill directs the agent to utilize a web search tool to retrieve contemporary facts during its analysis process. This is a functional requirement for the agent's purpose and does not involve the execution of untrusted scripts or the downloading of malicious payloads.
- [SAFE]: An indirect prompt injection surface is present as the skill processes untrusted user input and web search results.
- Ingestion points: User queries and results from the web search tool as defined in SKILL.md.
- Boundary markers: Not explicitly defined in the instructions for external data.
- Capability inventory: Web search functionality.
- Sanitization: Not specified. This profile is consistent with standard AI agents designed for real-time information retrieval.
Audit Metadata