The Agent Skills Directory

[SAFE]: The skill serves as a high-quality educational resource focused on LLM best practices, project planning, and architectural decisions.
[INDIRECT_PROMPT_INJECTION]: The prompt_eval_runner.py script ingests model output strings from external JSONL files for validation. This process is performed locally without execution, posing no threat to the environment.
[EXTERNAL_DOWNLOADS]: The skill contains URL references for pricing and documentation purposes, targeting well-known technology providers such as OpenAI, Anthropic, and Google. These are informational and originate from trusted sources.
[COMMAND_EXECUTION]: Utility scripts provided for cost estimation and evaluation use standard library functions only and do not incorporate user input into shell commands or unsafe execution blocks.

ai-llm