The Agent Skills Directory

[SAFE]: The skill is designed for educational and diagnostic purposes regarding LLM context performance. No malicious patterns or security vulnerabilities were identified in the analysis.
[SAFE]: The Python module scripts/degradation_detector.py uses standard numerical processing with numpy to simulate attention patterns. It lacks any network, file system, or dynamic code execution capabilities.
[SAFE]: No instances of prompt injection, data exfiltration, or obfuscation were found in the markdown files or code snippets.

context-degradation