Track LLM API costs and cache responses in Python apps.
npx skills add https://github.com/AbYousef739/clawcache-free --skill clawcache-free
Please help me install this Agent Skill. Skill: ClawCache Free Source: https://github.com/AbYousef739/clawcache-free/tree/main Command: npx skills add https://github.com/AbYousef739/clawcache-free --skill clawcache-free
ClawCache Free helps developers automatically track every penny spent on LLM API calls and cache responses to dramatically reduce costs, with per‑call token counting, daily reports, and multi‑provider support.
Install clawcache and wrap your LLM call with monitor_cost or async_monitor_cost to start tracking costs and caching.
High-intent search queries and answers about installing and using this skill.
Yes, ClawCache Free supports multi-provider cost tracking for OpenAI, Anthropic, Mistral, and Ollama. You can monitor per-call token counting and generate daily CLI reports showing spending and cache savings across all these providers simultaneously.
ClawCache Free uses SQLite with composite keys combining the model, prompts, and parameters for exact-match caching. It features WAL mode and cross-platform file locking to ensure cache security and reliability during concurrent Python application operations.
Yes, ClawCache Free supports token counting and cost tracking for local models using Ollama. It automatically logs per-call token usage and generates daily CLI reports showing both spending and cache efficiency for local LLM infrastructure.
ClawCache Free requires pickle to be explicitly enabled as an opt-in feature. This security measure ensures safe caching operations, while SQLite WAL mode and cross-platform file locking handle concurrent access without risking data integrity.
Yes, ClawCache Free works with both synchronous and asynchronous Python LLM calls. You simply wrap your LLM call with monitor_cost or async_monitor_cost to automatically start tracking costs and caching responses across your supported providers.
lguibrInstrument applications to generate LangSmith traces for LLM observability.
PostHogAggregate LLM spend and token metrics from PostHog events.
GGEduRoute LLM tasks to cost-appropriate models and track cumulative API spend.
BagelHoleImplement multi-layer LLM caching with Redis, GPTCache, Qdrant, and provider-side prompt caching.
