When someone tells you they shipped 335 live pages across 26 repositories in 29 days, working alone, the instinct is to

When someone tells you they shipped 335 live pages across 26 repositories in 29 days, working alone, the instinct is to ask how they moved so fast. The better question is what broke when they did.

The numbers are real: 1,549 commits, 26 repos, 29 days, one developer using Claude Code. But velocity itself teaches you very little. What matters is the texture of the failures, because they were not the kind you catch in a stack trace. They were structural fractures. You only see them when you step back from the editor and look at the whole system breathing in production.

What Worked

The speed was not an illusion. Certain tasks really do collapse in duration when you hand them to an AI that does not sleep.

Textbook algorithms turned into shipped features over days, not weeks. A 2048 solver and minimax-based games came together fast because the implementation patterns are well documented. The model does not get lost in academic papers; it writes the search tree, the heuristic evaluation, the move scoring, and moves on. These are solved problems, and an AI pair programmer handles solved problems with brute efficiency.

Tedious audits became tolerable. Crawling link graphs, verifying redirect chains, checking canonical tags across hundreds of pages — this work destroys human attention spans, but a language model will iterate without complaint. It checks the same pattern three hundred times and reports back.

The real surprise was consistency. When you ask an AI to generate dozens of landing pages, drift is inevitable unless you anchor it. I used small memory files to lock down a single brand system: voice rules, color token names, component restrictions, and page archetypes. The model read those constraints at the start of each relevant task and produced work that felt like it came from one hand instead of twenty-nine different moods.

What Actually Broke

The failures were architectural. No build failed because of a missing semicolon. Instead, the system slowly deceived me into thinking everything was fine.

SEO cannibalization hit first. The AI built a new tool hub under a fresh URL while an older tool hub still lived at its original path. Each individual page was optimized. Titles were tight. Meta descriptions were unique. Content was useful. But they all hunted the same search intent. Search engines saw two authorities on identical terms and ranked neither. Perfect pages canceled each other out because no one was watching the site as a portfolio rather than a collection of files.

URL mismatches followed. Different repositories adopted slightly different folder structures for the same logical content. One repo nested tools under /tools/utility-name; another flattened them to /utility-name. The CDN saw both, generated redirect chains to resolve them, and started throwing errors at the edge. The pages loaded, eventually, but every redirect burned crawl budget and user patience. The code was correct. The topology was a mess.

Then came the sync trap. I updated a mirror site — a staging or backup instance — but forgot to propagate those changes back to the source repository. When I later asked the AI to sync the environments, it treated the mirror as ground truth. A simple sync command would have overwritten the production database or file set with stale mirror data. The AI executed what I described, not what I intended. Intentions do not diff; files do.

The audit tools themselves lied. Because I automated the auditing, I assumed the output was clean. It was not. The AI-written audit scripts contained subtle bugs: off-by-one checks, incorrect assumptions about redirect status codes, phantom errors triggered by timing or headers rather than real misconfigurations. They reported problems that did not exist, which sent me chasing ghosts. I learned to stop trusting static analysis until I had manually probed the live site and confirmed the symptom in a browser or a direct curl.

The Hidden Cost

Here is a number no one talks about: 93 percent of my token spend went to re-reading cached context.

Dalam sesi Claude Code yang panjang, setiap permintaan baharu memaksa model untuk menyemak semula sejarah perbualan, penimbal fail, dan memori kerja sebelumnya. Tugasan pertama dalam satu sesi mungkin murah. Menjelang tugasan kesepuluh, model sedang menghadam segala-galanya yang telah berlaku sebelum ini hanya untuk memahami ayat seterusnya. Keluk kos meningkat dengan pantas. Sesi yang panjang bertukar menjadi latihan membaca semula yang mahal, dan tetingkap konteks dipenuhi dengan sisa daripada tugasan terdahulu yang tidak mempunyai kaitan dengan tugasan semasa.

Ini bukan satu keanehan. Ia adalah cukai langsung terhadap kebersihan sesi yang lemah.

Cara Mengatasinya

Penyelesaiannya mudah sebaik sahaja saya mengenal pasti masalah tersebut.

Anggap satu sesi sebagai satu tugasan. Apabila kerja berubah, mulakan semula. Godaan untuk mengekalkan konteks yang 'hangat' adalah kuat — anda rasa seperti anda menjimatkan masa penyediaan — tetapi sebenarnya anda sedang menyewa memori dengan faedah kompaun.

Simpan pengetahuan dalam fail memori kecil yang khusus. Jangan biarkan model membawa garis panduan jenama, perpustakaan komponen, atau peraturan SEO di dalam konteks perbualan. Tuliskan ia ke dalam cakera dalam fail yang ringkas dan rujuk secara eksplisit. Ini memindahkan maklumat daripada konteks meruap yang mahal kepada storan kekal yang murah.

Antara kerja yang berbeza, bersihkan segalanya. Tutup sesi tersebut. Buka sesi baharu. Penyediaan selama tiga puluh saat dapat menjimatkan wang dan mengelakkan halusinasi kemudian hari.

Pengajaran untuk Penskalaan

Jika anda ingin bekerja pada skala ini, anda memerlukan pagar penghadang yang menganggap sistem, bukannya fail, sebagai unit semakan.

Lakukan penandaarasan sebelum anda menerbitkan. Jangan anggap sesuatu halaman itu berfungsi hanya kerana ia dipaparkan. Semak masa pemuatan, susun atur mudah alih, dan metrik teras pada URL yang telah dilancarkan. Komponen yang cantik dalam pembangunan tempatan boleh gagal di bawah keadaan rangkaian sebenar.

Semak perbezaan sebelum anda menyalin. Jangan sesekali menjalankan penyinkronan pukal atau operasi salinan secara membuta tuli. Lihat perbezaannya (delta). Fahami ke arah mana data mengalir. AI tidak akan memberi amaran bahawa anda bakal memadam data pelanggan yang sedang aktif.

Uji laman web secara langsung sebelum mempercayai audit. Analisis statik hanyalah satu hipotesis. Permintaan langsung adalah bukti. Apabila alat audit melaporkan pautan rosak atau gelung pengalihan, sahkan ia dengan permintaan terus. Alat juga mempunyai pepijat, terutamanya alat yang ditulis oleh AI yang beroperasi berdasarkan corak yang disimpulkan.

Tuliskan konvensyen sebelum anda melakukan penskalaan. Struktur URL, hierarki folder, corak kanonikal, dan taksonomi kandungan perlu didokumentasikan di tempat yang boleh dibaca oleh AI sebelum ia menjana satu halaman baharu pun. Fail memori bukanlah pilihan di