The latest articles on DEV Community by Syed Akmal (@sayasyedakmal). https://dev.to/sayasyedakmal https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F718352%2F5862b818-ce4e-489f-92a0-6a78e9327850.jpg https://dev.to/sayasyedakmal en Syed Akmal Thu, 03 Sep 2026 04:51:26 +0000 https://dev.to/sayasyedakmal/decoding-the-new-ai-lingo-loops-harnesses-squads-hill-climbing-oh-my-3kpe https://dev.to/sayasyedakmal/decoding-the-new-ai-lingo-loops-harnesses-squads-hill-climbing-oh-my-3kpe <p><em>Summary of an article by Cassidy Williams, GitHub Blog (September 2, 2026), based on a GitHub Podcast episode.</em></p> <h2> Loop Engineering </h2> <p>Building repeatable systems around AI agents instead of manually prompting one task at a time.</p> <p>Example: a scheduled loop that fetches issues, has an agent summarize them, checks the output, and escalates problems — basically an "AI-native cron job."</p> <h2> Ralph Loops </h2> <p>A rougher version of loop engineering. You give an agent a detailed task (like a spec document) and let it keep working until done.</p> <ul> <li>Useful for breaking big tasks into repeated plan-act-check cycles</li> <li>Can burn a lot of tokens and compute, since every retry costs more</li> </ul> <h2> Squads and Fleets </h2> <ul> <li> <strong>Squad</strong>: a group of agents with different jobs (one plans, one checks the plan, one builds it, one tests, one reviews) — like a real team</li> <li> <strong>Fleet</strong>: multiple agents working in parallel</li> </ul> <p>You can run a squad within a fleet, either in parallel or in sequence. The point is splitting work by role instead of one agent doing everything.</p> <h2> Harnesses </h2> <p>Everything around the AI model that makes it actually useful: tools, permissions, memory, context, orchestration.</p> <p>Named after horse harnesses — the model is the horse's raw power, the harness directs it safely. GitHub Copilot is an example: it connects the model to codebases, editors, pull requests, and terminals.</p> <p>"Harness engineering" means building and improving that surrounding system.</p> <h2> Hill Climbing </h2> <p>The ongoing process of improving agents and harnesses using feedback and evals.</p> <p>Example: checking whether an agent reviewing pull requests actually finds real bugs, then adjusting the tooling to get better results.</p> <h2> Forward Deployed Engineer </h2> <p>Not really a new role — it's a customer-facing engineer (sales/solutions engineer) who helps clients integrate technical solutions. Now often focused specifically on AI tools and agents.</p> <h2> Closed Models vs. Open Weights vs. Open Source Models </h2> <div class="table-wrapper-paragraph"><table> <thead> <tr> <th>Type</th> <th>What's available</th> </tr> </thead> <tbody> <tr> <td>Closed models</td> <td>Only API/product access; no weights, data, or training process (most big frontier models)</td> </tr> <tr> <td>Open weight models</td> <td>Trained weights are downloadable and can run locally; training data/method may stay private</td> </tr> <tr> <td>Open source models</td> <td>Model, code, data, and training process are all available for inspection and modification</td> </tr> </tbody> </table></div> <h2> Takeaway </h2> <p>These terms will keep evolving — some will stick, some won't. What matters more than the buzzwords is the underlying practice: can your workflow repeat reliably, how do you validate output, how much human oversight is needed, and how do you keep improving the system.</p> <p><em>Source: <a href="https://github.blog/ai-and-ml/decoding-the-new-ai-lingo-loops-harnesses-squads-hill-climbing-oh-my/" rel="noopener noreferrer">github.blog</a></em></p> agents ai automation github