Pi, a coding agent from pi.dev, implements compaction to manage LLM context window limits by summarizing older conversation content while preserving recent messages. Compaction triggers automatically …
Pi, a minimal coding harness from earendil-works, outperformed Claude Code and Codex at lower cost in Databricks' benchmark of coding agents on its multi-million line codebase, and Shopify built its p…
Inference APIs are increasingly returning provider-bound state that makes AI sessions non-portable, according to a critical analysis by Earendil. Features like encrypted content and server-side storag…
Prompt caching, which reuses computed attention states from previous LLM requests, is critical for making coding agents economically viable but remains fragile to changes in tool definitions, model sw…