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.

Katika kikao kirefu cha Claude Code, kila ombi jipya hulazimisha modeli kupitia tena historia ya mazungumzo yaliyopita, file buffers, na kumbukumbu ya kazi (working memory). Kazi ya kwanza katika kikao inaweza kuwa na gharama nafuu. Kufikia kazi ya kumi, modeli inakuwa inatafuta na kuelewa kila kitu kilichotangulia ili tu kuelewa sentensi inayofuata. Mchirizi wa gharama unazidi kupanda kwa kasi. Vikao virefu vinageuka kuwa mazoezi ghali ya kusoma upya, na dirisha la muktadha (context window) hujaa takataka kutoka kwa kazi za awali ambazo hazina uhusiano wowote na kazi ya sasa.

Hili si jambo la ajabu tu. Ni kodi ya moja kwa moja kwa usafi mbaya wa kikao.

Jinsi ya Kurekebisha

Marekebisho yalikuwa rahisi mara tu nilipoyataja matatizo.

Chukulia kikao kimoja kama kazi moja. Kazi inapobadilika, anza upya. Tamaa ya kuweka muktadha ukiwa tayari ni kubwa — unahisi kama unaokoa muda wa maandalizi — lakini kwa kweli unakodisha kumbukumbu kwa riba inayoongezeka.

Weka maarifa katika faili ndogo za kumbukumbu maalum. Usiruhusu modeli ibebe miongozo ya chapa, maktaba za vipengele, au sheria za SEO ndani ya muktadha wa mazungumzo. Ziandike kwenye diski katika faili fupi na uzitaje waziwazi. Hii inahamisha habari kutoka kwenye muktadha ghali wa muda mfupi kwenda kwenye hifadhi ya kudumu na rahisi.

Kati ya kazi tofauti, safisha kila kitu. Funga kikao. Fungua kingine kipya. Sekunde themini za maandalizi huokoa madolara na upotoshaji (hallucinations) baadaye.

Mafunzo kwa ajili ya Kupanua

Ikiwa utafanya kazi kwa kiwango hiki, unahitaji vizuizi (guardrails) vinavyochukulia mfumo, na si faili, kama kitengo cha mapitio.

Fanya upimaji (Benchmark) kabla ya kuchapisha. Usichukulie kuwa ukurasa unafanya kazi kwa sababu tu unaonekana. Angalia muda wa kupakia, mpangilio wa simu, na vipimo muhimu kwenye URL iliyowekwa. Kipengele kizuri katika maendeleo ya ndani kinaweza kufeli chini ya hali halisi ya mtandao.

Linganisha (Diff) kabla ya kunakili. Usifanye operesheni ya usawazishaji wa wingi au unakili bila kuangalia. Angalia tofauti (delta). Elewa upande ambao data inatiririka. AI haitakuonya kwamba unakaribia kufuta na kuandika upya data halisi ya mteja.

Chunguza tovuti zinazoishi kabla ya kuamini ukaguzi. Uchambuzi wa tuli ni nadharia tu. Ombi la moja kwa moja ni ushahidi. Wakati zana ya ukaguzi inaripoti kiungo kilichovunjika au mzunguko wa uelekezaji (redirect loop), kithibitishie kwa ombi la moja kwa moja. Zana pia zina hitilafu, hasa zana zilizoandikwa na AI inayofanya kazi kwa kutegemea mifumo iliyodhaniwa.

Andika kanuni kabla ya kutanua. Muundo wa URL, mpangilio wa folda, mifumo ya kanonika, na taksonomia ya maudhui yanahitaji kuandikwa mahali ambapo AI inaweza kusoma kabla haijatengeneza ukurasa mmoja mpya. Faili za kumbukumbu si hiari katika