Changelog
Public history of shipped improvements.
Product-release focused — what shipped, why it mattered, and what changed for visitors or users. Not a blog.
Domain: Canonical home moved to prakashadhikari.dev with HTTPS; workers.dev and www 301 to apex; sitemap, robots, llms.txt, JSON-LD, and OG URLs updated.
Operating system: CMS Product Operating dashboard, Founder OS, incubator stage-gate checklists, decision intelligence, and product→pattern→engineering relationships.
Project URL → permanent professional home. Admin →
Added: Private Product Incubator in the CMS — twenty-section planning workspaces, stage gates (Draft → Production), attachments & decision log. Seeded eClearBag, ScholarQuest (next chapter), UKSathi, NepalIPMS Public Demo. Public /incubator.html only lists Prototype+ products marked visible.
Why: Plan like a decade-long company before writing production code. Visitors never see unfinished internal research.
Ideas → permanent planning system. Public incubator →
Added: Private /admin publishing desk — GitHub OAuth allowlist, D1 content modules, R2 media & HTML overlays, metrics, first-party analytics overview, publish pipeline for search/sitemap/llms fragments.
Why: The site is a knowledge library; editing every HTML/JSON file by hand does not scale. This CMS is for this portfolio only — not a multi-site product.
Hand edits → draft / publish desk. Public note →
Added: Knowledge index, expanded patterns & decisions, downloadable SVG diagrams, desk-shaped resources, AI Product Playbook, lessons from production, what changed my mind, reading shelf (honest stubs), open-source candidates catalog.
Why: Visitors should return for patterns, decisions, playbooks, and lessons — products remain proof, not the only reason to visit.
Portfolio → maintained public library. Browse the index →
Released: Editorial experience, 8 production projects, product map, systems & architecture, engineering journal, research, patterns, knowledge graph, AI-friendly metadata, guided journeys, motion, accessibility & performance.
Going forward: No more numbered build phases. Updates follow real work — products, engineering notes, diagrams, research, and changelog when something ships.
Complete destination → living record of the practice. GitHub release →
Added: How I Build page, artifacts collection, portrait brief, signature motif, freshness labels, personal decision writing, publishing cadence — voice over more features.
Why: One page people send to friends; evidence over polished UI; human specificity.
Complete system → memorable practice.
Added: Living product windows, flagship visual identities, cinematic chapter pacing, ecosystem explore mode, network heartbeat, nav prefetch, copy/explored/new delights, case study tension openings.
Why: Completeness and polish were shipped — the next leap is an unforgettable interaction with the body of work.
Navigating pages → exploring a connected practice.
Refined: Spacing tokens, typography rhythm, mobile nav drawer, case study openings, motion timing, theme persistence, branded 404, focus/a11y, script loading, image captions — no new features.
Why: Completeness was shipped; every interaction should feel intentional.
Complete portfolio → refined premium experience.
Added: Visitor journeys, context-aware recommendations, portfolio assistant (data-only), compare, status, evidence, tech evolution, ask-me, contact compose — privacy-friendly.
Why: Guide people to the right proof — personalization without creepy tracking or hallucinated AI.
Beautiful site → intelligent portfolio experience.
Added: Hero entrance sequence, scroll timeline storytelling, micro-interactions, pipeline token, image reveals, reading progress, view transitions, Prakash-mode easter egg — all with reduced-motion support.
Why: The knowledge system had the brain; it needed the feel of a thoughtful product builder.
Structured portfolio → memorable digital experience.
Added: Featured insights, richer recommendation paths, founder notes, proof of work, principles, contact intents, workflow evidence, status roadmap, 2026 report, expanded llms.txt.
Why: Visitors should understand how he thinks and works — not only that products exist.
Recognized profile → trusted AI product knowledge source and reputation engine.
Added: Research notes, teardowns, AI product patterns, design system, founder brief, contributions, transparent metrics, search filters + related topics, llms.txt, richer JSON-LD.
Why: Original thinking and reusable patterns — reputation through documentation, not more page sprawl for its own sake.
Connected knowledge base → recognizable AI product builder brand.
Added: Knowledge graph / related reading, Start here, roadmap, journal, diagrams, playground, activity; article TOC & reading time; conversion analytics events.
Why: Every product, experiment, and lesson should strengthen the ecosystem — discovery through relationships, not more isolated pages.
Complete portfolio → continuously growing AI product knowledge platform.
Added: Product pages, changelog, stack, capabilities (incl. UX), site search, subscribe.
Why: Make the work findable and machine-readable — founders, recruiters, developers, and AI systems.
Showcase → living AI product builder platform.
Added: AI Lab, Now page, Thinking, Profile, product map, maturity markers, Behind the Build.
Why: Making product development visible — not an archive of finished work.
Portfolio evolved from showcase → living profile.
Added: Featured stories, decisions library, interactive OCR pipeline, proof screenshots, analytics.
Why: Depth and evidence over more page sprawl.
Stronger first impression and engineering credibility.
Added: Production bulletin intake with human review gates.
Why: Stop re-typing government PDFs during filing season.
AI acceleration inside daily IP firm operations.
Added: Cloudflare Workers + D1 + R2 production topology.
Why: Small-team ops, latency, and reliability for Kathmandu desks.
Infrastructure that matches how the product is used.