the thing nobody mentions about long AI coding sessions

<!-- SC_OFF --><div class="md"><p>How do you guys deal with long context, for example while coding, when you’re going back and forth for adjustments or fixing some errors and

<!-- SC_OFF --><div class="md"><p>I organize everything by topic (work, study, documents, buying, recipes, etc.), and each topic stays in one long chat. Over time, those chats

<!-- SC_OFF --><div class="md"><p><strong>quick context:</strong> we're a small AI creative agency, 9 people, based in Pune. we work with brands on commercial visuals product

<!-- SC_OFF --><div class="md"><p>It’s been somewhere between 1 and 1.5 years since I last wrote a line of code.</p> <p>I wrote everything from Assembly and C to Python and

<!-- SC_OFF --><div class="md"><p>Cursor dropped their new product yesterday and its not a code editor anymore. its an agent orchestration console that happens to have an editor

<!-- SC_OFF --><div class="md"><p>The productivity metrics on AI coding tools focus almost entirely on acceptance rate and time saved. The metric nobody is tracking is technical

<!-- SC_OFF --><div class="md"><p>Heptabase just dropped a CLI so Claude Code / Codex can create, read, and update a local knowledge base from the terminal. It’s a smart move.</p>

<!-- SC_OFF --><div class="md"><p>i’ve been playing around with coding agents recently and kept running into the same issue:</p> <p>they get stuck in loops</p> <pre><code>fail →