LLM orchestration

Open-source local AI coworker system that routes Claude Code and Codex work to local models on your own hardware, with durable task handling, multi-machine cooking, and MCP/CLI support.

A Python MCP server that lets a parent LLM delegate work to named sub-agents defined in YAML, with per-agent prompts, model settings, and downstream MCP tool access.

A Python-based coordination protocol and daemon for running multiple LLM sessions in parallel with shared awareness, aimed at Claude Code and other Anthropic-compatible models.

A catalog of design patterns for building agentic systems that verify outputs with evidence instead of relying on self-review. The page explains two main failure modes and presents structured patterns for context, verification, and orchestration.