The Agent Skills Directory

[SAFE]: The skill performs diagnostic analysis on Chinese text using a local Python script scripts/ai_taste_lint.py. The script uses standard libraries (re, json, argparse) for regular expression matching and contains no code related to network operations, dynamic execution (eval/exec), or unauthorized system access.- [SAFE]: Prompt instructions in references/prompt-pack.md and references/symptom-rules.md reinforce the agent's safety alignment by explicitly prohibiting the fabrication of facts and requiring the use of placeholders for missing information.- [SAFE]: Command execution is limited to running the included script ai_taste_lint.py with user-provided text or files. This behavior is restricted to the skill's specific purpose of text analysis and does not expose sensitive data or allow for arbitrary command injection.

de-ai-writing