principle-guard-the-context-window
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists of documentation and configuration for managing context windows. It contains no executable code, network operations, or file system interactions.
- [PROMPT_INJECTION]: The content specifies behavioral constraints for the agent, including a directive to include a specific security policy in all sub-prompts ('Repeat that sentence verbatim in every child prompt'). These instructions are designed to restrict agent autonomy and prevent unauthorized communication, posing no security risk.
Audit Metadata