clip-aware-embeddings

This skill automatically intercepts and optimizes complex prompts using a prompt-learning MCP server, applying techniques like APE and DSPy. It maintains an embedding-indexed history to learn from past performance and improve over time. Use it when a user requests prompt optimization or is tackling a complex task, but not for simple questions or direct commands.

This skill provides expert guidance for building production-grade chat bots on platforms like Discord, Telegram, and Slack. It covers bot-specific architecture including rate limiting, state machines, event sourcing, and moderation systems. Use it for chat automation projects, not for general web APIs or frontend components.

This Claude Skill automatically intercepts and optimizes complex prompts using a prompt-learning MCP server, applying techniques like APE and DSPy. It learns from historical performance via embedding-indexed feedback to improve prompts over time. Use it for explicit optimization requests or complex tasks, but not for simple questions or direct commands.

This skill enables semantic image-text matching using CLIP models for tasks like image search, similarity matching, and zero-shot classification. It provides embeddings that align visual and textual concepts but excludes fine-grained classification, object counting, and spatial reasoning. Developers should activate it with terms like "CLIP," "embeddings," or "semantic search" when working with visual-semantic alignment tasks.