Pi Durable is a lightweight, malleable harness for building long-running, durable agentic applications that complements Pi the coding agent rather than replacing it. It bundles storage backends (memory, SQLite, JSONL) with a small, portable storage interface that can run across JavaScript runtimes (Node, Bun, Cloudflare Durable Objects) and execution environments (Node local files or remote VMs). The codebase is intentionally compact - about 15,000 lines - and keeps only the active working set in memory: active transcripts, live tasks, and pending submissions, while compacting older messages so conversations with tens of thousands of messages remain bounded by model context windows.
Durability and concurrency are core: every step is a checkpointed task so runs survive crashes, restarts, and migrations; a requestId ensures submissions are exactly-once. One harness runs many concurrent conversations and forks (think a Slack channel and threads), each with its own agent configuration and working directory. Extensible bundles provide system prompt sections, tools, hooks and durable tasks; prompt sections are rebuilt per request and recorded in transcripts so restarts/forks see the same view. Tools declare replay semantics (safe to rerun or not), enabling safe automatic retries for read-only actions and explicit handling for side-effectful operations like deploys.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
AstaBrief is an open-source model designed for fast scientific report generation, capable of turning research questions and literature into cited reports. It reduces generation time by nearly three and a half times compared to proprietary models, making it a useful tool for researchers. (allenai.org)
Shrivu Shankar argues that SaaS companies are evolving into 'harnesses,' where core tasks are performed by AI agents rather than humans. This shift transforms organizational structure, with humans overseeing and reviewing AI-driven processes within the company. (blog.sshh.io)
Three AI agents - Meta’s Muse, Anthropic’s Claude, and OpenAI’s GPT - were tested on multilingual research tasks involving U.S. and Iranian data. The experiment focused on how language and context influence agent performance across reasoning, sourcing, and artifact creation. (royapakzad.substack.com)
An open-source project named LDRAW-NOVA allows users to generate Lego models using AI. The repository contains tools, examples, and documentation for creating Lego designs with artificial intelligence. (github.com)
DwarfStar 4 (ds4) is a local inference engine that runs large language and vision models on high-memory Mac, CUDA, and ROCm machines. It supports models like DeepSeek V4 and Qwen3.8, offering CLI, APIs, and persistent cache features for efficient local deployment. (dwarfstar.sh)
Rai is a CPU-only large language model inference engine written entirely in Rust. It aims to provide efficient and pure Rust solutions for LLM deployment without relying on GPUs. (github.com)
Today's best Hacker News stories, summarized and screenshotted, one email a day.