Tag
All content about Harness, organized for fast scanning.
Recent developments in AI coding agents highlight a shift towards collaborative and persistent coding environments, with tools designed to enhance teamwork and efficiency. Innovations like DeepSeek Harness and Meta's Muse Code are entering beta phases, focusing on integrating long-running agents and improving code management. Experts emphasize the importance of human oversight and structured constraints to ensure code quality and effective implementation in these evolving systems.
OpenClaw 2.0 aims to shift teams from local harnesses to a shared coding agent that tracks work and coordinates tasks. Peter Steinberger says it can run across connected machines and cloud sessions, with pairing, WebVNC testing, and takeover support.
Addy Osmani says automated code generation still needs humans upstream, where intent, architecture, and quality bars are set. His “software factory” model emphasizes deterministic checks for evidence, with reviews focused on high-risk changes and clear ownership.
Meta has just rolled out Muse Code in beta, bringing persistent background agents to tackle large codebases from the terminal. Powered by Muse Spark 1.2, it supports parallel sub-agents and a local event log to resume work after crashes.
DeepSeek Harness has entered a closed beta, with invites aimed at developers building open-source projects tied to Agent Harness. Jiayuan Zhang says interested builders should reply with a GitHub ID and representative project details.
Addy Osmani says code quality in the agent era hinges less on what humans can review and more on the constraints around the system. His proposed “exit gate” approach layers checks for security, performance, cost, and more before anything ships.
Addy Osmani says coding agents may be shifting from prompt-by-prompt use to goal-driven loops that plan, split work, verify results, and repeat. His thread maps the core building blocks—and the token, security, and “exit condition” pitfalls. [https://x.com/addyosmani/status/2064127981161959567](htt…
Anthropic has just rolled out Defending Code Reference Harness, an open-source reference for using Claude to find and fix vulnerabilities. It includes guided Claude Code skills plus a sandboxed C/C++ pipeline with ASAN and gVisor, though the repo is unmaintained.
Basis shared how it reorganized its monorepo to better support coding agents, built around “canonical” context and a six-layer AGENTS.md and skills architecture. The company claims the changes drove 5x higher token usage per developer and 2.5x faster weekly commits.
Anshuman Mishra lays out a bottom-up recipe for agent training using a tiny text-to-diagram task. The key: start with a strict environment and reward loop, use SFT to learn valid actions, then apply RL to optimize behavior—and watch for reward hacking.
A recent article by Addy Osmani explores “long-running agents” that can keep working across sessions without losing state. It outlines the key architectural patterns and why today’s agents still stall when context, memory, and verification break down.
Joongwon Kim and coauthors argue test-time scaling for long-horizon coding agents depends less on more sampling and more on carrying forward useful rollout information. Their summary-based RTV and PDR methods boost results on SWE-Bench Verified and Terminal-Bench v2.0.