# Awesome Harness Engineering

A curated list of articles, playbooks, benchmarks, specifications, and open-source projects for harness engineering: shaping the environment around AI agents so they work reliably.

Awesome Harness Engineering is a curated list of resources for harness engineering, defined as the practice of shaping the environment around AI agents so they can work reliably. It sits at the intersection of context engineering, evaluation, observability, orchestration, safe autonomy, and software architecture, and focuses on resources that make agents more dependable in long-running coding and research workflows; generic agent tooling is out of scope unless it directly covers harness design, context management, evaluation, runtime control, or other reliability-critical primitives.

Its sections cover courses and learning resources, context and memory design, constraints and guardrails including tool execution boundaries and security and authorization, specs and agent instruction file formats, evaluation and observability including benchmarks for coding, terminal, web, GUI, and multi-agent tasks, and runtimes and reference implementations spanning sandboxes, coding-agent harnesses, multi-agent orchestration, and browser and MCP tool integration.

Website: https://github.com/walkinglabs/awesome-harness-engineering