The Agent Skills Directory

[SAFE]: The skill implements deterministic keyword-based routing via scripts/classifier.py rather than relying on LLM reasoning for orchestration, which prevents query manipulation from affecting control flow.
[SAFE]: The Python scripts for classification, decomposition, and logging use only standard libraries and perform safe operations within the user's home directory.
[SAFE]: Instructions include rigorous Agent Integrity Rules such as source discipline and mandatory audit logs, which mitigate hallucination and data fabrication risks.
[SAFE]: No obfuscation, hardcoded credentials, or dangerous remote code execution patterns were found in the skill's instructions or scripts.

research