Description

An experimental harness designed to improve the reliability of local AI models when used as coding agents in Pi. It provides specialized safety controls, such as loop detection and memory management, to prevent common local LLM failure patterns like repeated tool calls or cut-off edits. The repository includes specific adapters for DeepSeek, Qwen, Laguna, and Ornith models.

Descriptions, tags, and model credits may be AI-generated or inferred from public sources and can be incomplete or wrong. Learn more.