AGENT-ARCHITECTURE

How a one-developer protest against bloated coding harnesses became a 95,000-star agent toolkit: pi's five-package architecture, branching JSONL session trees, four run modes, and the philosophy that refuses to build sub-agents, plan mode, or MCP.

Agents that critique their own output, learn from mistakes, and get better over time - the three patterns that actually ship, from simple reflection loops to tree search and meta agents.

The 2026 agent decision is not CrewAI vs LangGraph. It is whether your loop lives in vendor infrastructure, a self-hosted graph runtime, or a plain while-loop you wrote yourself. Here is how to choose.

The Multi-Stream LLMs paper argues that agents are bottlenecked by single chat streams. The practical takeaway is not to rebuild everything today, but to design agent runtimes around separated channels.