https://meshrefine.com/en/tags/harness/Posts · Mesh RefinementMon, 24 Aug 2026 20:03:30 +0200https://meshrefine.com/en/posts/chronicles-aug-17-aug-23-2026/<p>Last week was all about the harness. Gone are the times when the models were all things-in-themselves. Now, as with early Homo erectus, the tools are the factor of survival. If we look at the types of harness-related posts and announcements, we will see three trends.</p> <h2 id="04602056b054f8f419959cfb26ab2e8f-harness-products">Harness products</h2> <p>Several companies have released either agents or infra for agents at once.</p> <p>OpenAI has <a href="https://developers.openai.com/blog/codex-as-a-platform">released the execution framework</a> that their Codex (the CLI one) is based on. Unsurprisingly called Harness, it is their answer to the <a href="https://code.claude.com/docs/en/agent-sdk/overview">Claude Agent SDK</a> and the <a href="https://docs.github.com/en/copilot/how-tos/copilot-sdk">GitHub Copilot SDK</a>. The Harness provides the execution loop, memory, tools, and other necessities of agentic life. You can use it in three different ways. First, you can just use <code>codex exec</code> to run non-interactive jobs. Second, you can use the Codex SDK for building workflows. And last, you can use app-server to build apps that require conversation handling. One point worthy of attention is that they claim that the Harness improved the performance of the GPT-5.6 Sol model from 13.3% to 38.3% on ARC-AGI-3. It&rsquo;s unclear what harness (not capitalized) was used as a baseline, though.</p>Mon, 24 Aug 2026 20:03:30 +0200https://meshrefine.com/en/posts/chronicles-aug-17-aug-23-2026/