World Wire

Microsoft Research has open-sourced Agensh , a decentralized harness for large teams of coding agents. The Agensh multi-agent harness removes the central orchestrator that most agent frameworks depend on. Every worker picks its own sub-task, shares findings and merges code into a common Git repo. On ProgramBench, the team scaled the setup from 1 agent to 1,024 agents. TL;DR A Harness : Tested from 1 to 1,024 agents. Each worker gets a 272,000-token input and 128,000-token output limit. Runs on: Linux with Docker Compose. Single node by default (8 workers). The 1,024-agent run used 16 nodes, 64 agents each. Workers have a 64 GiB memory ceiling. Performance: more agents raised scores on all 5 hardest ProgramBench tasks, and reached them sooner. Best: 79.59% pandoc pass rate with 32 agents on GPT-6.1-sol. GPT-5.6-sol hit 55.06% with 1,024 agents. Worst: the 5-task mean reached only 28.78% at 128 agents. Going from 1 to 8 agents added 1.37 points. What is...

In this tutorial, we work with Underdog Saluki 27B , a 7.9 GB build of Qwen3.8-27B that its makers say keeps tool calling intact at about two bits per weight and even edges out the full 54 GB model on their own 120-task benchmark. Saluki is built on ISTA-DASLab’s GSQ-RCO quantization of the same model, so we put both files side by side in stock llama.cpp. We start with the files themselves: a tensor-by-tensor diff of the two GGUF headers, then a byte-level comparison that shows whether Saluki is a remix of ISTA’s bits or a changed model. We then serve each file on one GPU, read its memory and speed, and score both on the same seeded 120 tasks from the Berkeley Function Calling Leaderboard, using BFCL’s own checker with strict and lenient parsing. Copy Code Copied Use a different Browser import os import re import ast import sys import json import math import copy import enum import glob import time import types import random import shutil import struct import hash...

In this tutorial, we build a stateful engineering copilot with Letta that demonstrates how agentic memory extends beyond conventional prompt-based conversations. We begin by creating inspectable core memory blocks that the agent can autonomously update, then expose those blocks for programmatic state synchronization and shared access across multiple agents. We further extend the architecture with archival vector memory for semantic retrieval, a custom server-side tool for deterministic risk estimation, streaming for observing agent execution, server-side message history for persistent conversations, and .af export for portability. Together, these components demonstrate a practical memory architecture in which persistent state, retrieval, tools, multi-agent collaboration, and execution traces work as an integrated agentic system. Copy Code Copied Use a different Browser import subprocess, sys subprocess.run([sys.executable, "-m", "pip", "install", ...