How adityaprakash-work/contextcraft works
Project type and purpose
• This is a Python library (distributed via PyPI / Poetry) called “contextcraft” whose goal is “Advanced context comprehension and manipulation for LLMs.” It provides abstractions for building up context, calling one or more LLM back-ends, post-processing, and persisting results.
File structure
• Top‐level files:
• .env.example – environment variable template for API keys, etc.
• pyproject.toml – declares Python dependencies (e.g. OpenAI).
• media/files – a placeholder for user-provided assets (text snippets, documents).
• output – where the tool writes generated results.
• scripts – any helper scripts (e.g. demo or bootstrap).
• src/contextcraft:
• config.py – loads and centralizes configuration (API keys, timeouts).
• lms/openai.py – adapter for interacting with the OpenAI API.
• programs/ – entry points or higher-level routines that glue config, context builders, and LLM calls.
Be detailed but not exhaustive
• Every module folder (config, lms, programs).
• Distinguish between what runs locally and what is an external dependency.
• How environment configuration (.env) flows into the system at startup.