The Agent Skills Directory

[SAFE]: The skill implements standard best practices for LLM cost optimization, including model routing based on task complexity and immutable cost tracking to prevent state corruption.
[SAFE]: It utilizes the official anthropic library for API interactions, which is a well-known and trusted package.
[SAFE]: The retry logic is appropriately scoped to transient errors (like network timeouts or rate limits), avoiding infinite loops or wasted budget on authentication failures.
[SAFE]: Prompt caching implementation follows official provider documentation to reduce token usage and improve performance for repetitive system prompts.
[SAFE]: No obfuscation, data exfiltration, or persistence mechanisms were detected in the provided code snippets or metadata.

cost-aware-llm-pipeline