boost-modules

What problem does it solve?

This Skill unit addresses the need for custom module creation within Harbor Boost, an optimizing LLM proxy. It simplifies the process of building Python modules that intercept or transform LLM chat completions, enabling advanced processing like reasoning chains, prompt injection, structured outputs, artifacts, and custom workflows.

Core Features & Use Cases

• Custom Module Creation: Develop Python modules to intercept and transform LLM chat completions.
• LLM Behavior Extension: Extend LLM behavior via proxy and implement chat completion middleware.
• Use Case: For a developer looking to integrate custom logic into their LLM interactions, this Skill unit allows for the creation of modules that can modify the output of LLM responses based on specific criteria.

Quick Start

To create a custom module for Harbor Boost, write a Python file with the structure defined in the SKILL.md file and place it in the 'custom_modules' directory. Restart the container and test the module through the API.