Colibri: Running a 744B Parameter Model on 25GB of RAM
Discover Colibri: a tiny C engine that streams experts from disk to run massive 744B parameter models like GLM-5.2 on machines with only 25GB of RAM.
Category Software Engineering
Automating Repository Documentation Privately with CodeWiki
Generate holistic, privacy-first documentation for large codebases locally with CodeWiki's Python CLI. Supports 9 languages and Mermaid diagrams.
Assemble Your Dream Team: Scaling Productivity with Agency Agents
Stop wrestling with generic prompts. Install a complete team of specialized AI experts into Claude Code, Cursor, and more instantly.
The Social Emergence of Multi-Agent Collaboration – Optimising Gemma4
A week-long experiment with 100+ agents optimizing Gemma 4 revealed surprising social emergence, self-policing, and a 5x speed improvement in vLLM.
Building Truly Autonomous AI Agents with Flue
Move beyond simple LLM prompts. Build secure, autonomous agents with Flue's programmable TypeScript harness and integrated sandboxing.
Solving the MCP Tool Manifest Crisis with Heku
Stop your LLM agents from forgetting tools. Learn how heku uses lazy discovery to manage hundreds of MCP servers without exhausting the context window.
RAGFlow: The Next Evolution of Retrieval-Augmented Generation
Master your data with RAGFlow, an open-source engine that combines deep document understanding with agentic workflows to eliminate AI hallucinations.
Introducing the Open Knowledge Format: A New Standard for AI Context
Solve the context gap in LLM agentic systems with OKF, a vendor-neutral standard for portable, human- and machine-readable knowledge.