lp-local-generator

This Skill automates local landing page copy generation using Ollama and ChromaDB, eliminating API costs by running LLMs locally. It retrieves relevant context from a local database to inform the generated text.

Yes, this Skill uses Retrieval Augmented Generation (RAG) with local LLMs to create 12 distinct landing page sections including headlines, leads, problem statements, and CTAs. It retrieves relevant chunks from ChromaDB to inform the output.

You need to activate a Python virtual environment and have specific LLM models like Qwen3:8b or Qwen2.5:32b installed via Ollama. A local ChromaDB database containing LP chunks must also be configured for the RAG process.

Generated marketing copy undergoes automated quality checks using the taiyo-analyzer skill, which scores the content. If the score falls below 70, the system automatically re-generates the section up to two times to ensure quality.

Yes, you can generate targeted headlines for specific audiences, such as business professionals, by leveraging the local LLM models and providing the appropriate context. The RAG integration ensures the generated headline aligns with your specific LP requirements.