2 added, 2 removed. Audit C to C.
# LLM Chat Skill
A robust support utility for interacting with Large Language Models (LLMs) via LiteLLM.
## Setup
1. All runtime dependencies (`typer`, `python-dotenv`, `rich`, `litellm`) are installed in `the Conda environment ` by `{{REPO_ROOT}}/install_skills.sh`.
2. The `.env` file at `{{REPO_ROOT}}/.env` must define:
- `LLM_MODEL`: The default text-generation model to use (e.g., `gemini/gemini-1.5-pro`, `ollama/llama3`).
- `LLM_BASE_URL`: Optional endpoint base URL. Use `http://localhost:11434` for local Ollama.
- `LLM_API_KEY`: Optional endpoint API key. Leave blank for local Ollama.
- `GEMINI_API_KEY`: Your Google Gemini API key.
- `OLLAMA_HOST`: The host URL for Ollama (if using Ollama).
- - `OLLAMA_NUM_CTX`: The baseline token context for Ollama.
+ - `OLLAMA_CONTEXT_LENGTH`: The baseline token context for Ollama.
## Usage
`llm_chat` is not exposed as a harness slash command. User-facing harness commands
call it internally through the shared skill APIs. Run it directly only when testing
provider configuration or debugging model behavior.
```bash
# Basic usage with default model
conda run -n sictic-env python -m skills.llm_chat "What is startup due diligence?"
# Override the default model
conda run -n sictic-env python -m skills.llm_chat "Summarize the risks." --model ollama/llama3
```
## Features
- **LiteLLM Integration:** Unified interface for hitting different providers (Gemini, Ollama, etc.).
- - **Ollama Context Management:** Automatically sizes the prompt context between `OLLAMA_NUM_CTX` and `OLLAMA_NUM_CTX_MAX`.
+ - **Ollama Context Management:** Automatically sizes the prompt context between `OLLAMA_CONTEXT_LENGTH` and `OLLAMA_CONTEXT_LENGTH_MAX`.
- **Rich Formatting:** Console outputs and markdown are beautifully rendered using the `Rich` library.
- **Robust Error Handling:** Catches connection errors and missing model definitions gracefully.