{hjemmesidekongen}/ai

Roadmap

Where hjemmesidekongen/ai is headed — upcoming features, new plugins, and the Blik desktop app.

Roadmap

hjemmesidekongen/ai is a plugin ecosystem that turns Claude Code into a full-service development environment — from brand identity through deployed applications. Here's what's coming next.

Priorities shift as we build. This page reflects current direction, not commitments. Features may ship in a different order or form than described here.

At a Glance


Blik — Desktop Mission Planner

In development

A native desktop application for planning, orchestrating, and monitoring your projects. Blik gives you a visual interface for the workflows that currently live in the terminal — project setup, pipeline execution, agent monitoring, and build progress — all in one window.


Brand Identity

Planned

A guided pipeline that produces a complete, machine-readable brand system from a short interview.

  • Identity interview — name, tagline, mission, values, positioning, and audience personas distilled into structured brand data
  • Visual system — color palettes with full tint/shade scales, typographic hierarchy, spacing system, and dark mode variants, all validated for WCAG AA accessibility
  • Logo generation — multiple design directions refined through feedback, exported as SVG variants (primary, reversed, monochrome, icon, horizontal, stacked) with clear space rules and misuse examples
  • Voice and tone — personality spectrum across formality, humor, enthusiasm, and technicality, with channel-specific variations and a vocabulary guide
  • Social media kit — Open Graph, Twitter, and LinkedIn templates matching your brand, plus content pillars and hashtag strategy
  • Brand manual — everything compiled into a polished document (Markdown and DOCX) with an interactive HTML preview

Brands can also be reverse-engineered from existing websites, documents, and social profiles.


Design System

Planned

End-to-end design pipeline that takes your brand and produces implementation-ready specs.

  • Design tokens — Tailwind config, CSS custom properties, and DTCG JSON generated from your brand, with a full accessibility contrast matrix
  • Creative direction — per-project personality layer defining motion philosophy, spatial rules, color strategy, typography personality, hero approach, and project-specific anti-patterns
  • Component library — structured specs for every interactive element (buttons, forms, cards, navigation), each tied to design tokens with accessibility annotations, responsive rules, and keyboard navigation specs
  • Page layouts — site structure, navigation maps, and page compositions built from your component library across breakpoints
  • Visual mockups — live Pencil designs with AI-generated imagery, exported as screenshots that serve as ground truth for implementation
  • Implementation guides — per-page motion choreography, entrance animations, hover behaviors, scroll interactions, and reduced-motion fallbacks, written for Motion.dev

SEO

Planned

Data-driven SEO strategy generation tied to your specific business, keywords, and competitors.

  • Keyword research — 30+ keywords organized by tier and search intent (informational, navigational, commercial, transactional) with volume and difficulty estimates
  • Competitor analysis — 3-5 competitors mapped by ranking keywords, domain authority, backlink profiles, and content gaps scored by opportunity
  • On-page rules — title patterns, meta description templates, heading hierarchy, internal linking strategy, and schema markup recommendations for your industry
  • Technical checklist — Core Web Vitals targets, mobile requirements, crawlability, and security items prioritized by severity
  • Content strategy — topic clusters with pillar pages and supporting content, a 3-month content calendar, and content type definitions
  • Link building — outreach targets, promotion tactics, and effort/impact estimates based on competitor backlink patterns

Content and Copy

Planned

Structured content generation that fills every slot in your design system with brand-aligned copy.

  • Page copy — headlines, body text, and CTAs for every page, written in your brand voice and organized by component slot
  • UX microcopy — error messages, validation feedback, tooltips, confirmation dialogs, loading states, and form labels, all following a consistent taxonomy
  • Content governance — writing standards, terminology guides, and review checklists that keep copy consistent as your team grows

Build Pipeline

Next

Multi-agent development execution that turns designs into working applications.

  • Feature decomposition — architectural analysis breaks features into components with clear file ownership boundaries
  • Wave planning — tasks organized into parallel and sequential waves, with model tier assignment matching complexity to capability
  • Parallel agent execution — specialist agents (frontend, backend, design-UX, documentation, devops) working simultaneously on non-conflicting files
  • Visual verification — automated screenshot comparison against Pencil mockups at section milestones, checking structural alignment rather than pixel perfection
  • Quality gates — security audit (OWASP top 10), code quality checks, design token compliance, test coverage validation, and spec alignment scoring before anything ships

DevOps

Planned

Deployment configuration and execution from a single command.

  • Provider detection — framework-aware recommendations (Vercel for Next.js, Railway for Node.js, Docker for anything)
  • CI/CD generation — GitHub Actions workflows covering lint, test, build, and deploy
  • Observability — error logging (Sentry), analytics instrumentation (PostHog/Mixpanel) with GDPR consent gating, and server-side health monitoring (OpenTelemetry) for projects with a backend
  • One-command deploy — build verification, provider deployment, and health check polling with automatic retry

Smarter Memory

Next

Improvements to how the system remembers and learns across sessions.

  • Episodic memory — long-term recall of what was tried, what worked, and what failed across projects
  • Project memory — temporal decay, two-tier storage, and deduplication so memory stays relevant without manual pruning
  • Continuous learning — instinct extraction from session observations, with qualified instincts graduating into skills, rules, or memory entries

Recently Shipped

A selection of what landed in the latest release cycle.

  • Context compaction — smart session recovery that preserves working state, active plans, and project context across /compact and session restarts
  • Planning engine — wave-based execution plans with dependency resolution, parallel dispatch, and verification gates between waves
  • 53 development skills — framework patterns (React, Next.js, Vue, Nuxt, Tailwind v4, NestJS, Prisma), testing disciplines (TDD, Playwright, Storybook), and Expo mobile development
  • Task management — Jira ingestion, local task storage, contradiction detection, QA handover, and PR workflows with Bitbucket integration
  • Workspace integrations — Jira, Confluence, Bitbucket, Slack, and Outlook via MCP, with pre-built configuration templates
  • Creator toolkit — skill, hook, command, agent, plugin, and MCP server creators with eval infrastructure and A/B testing
  • Code review — parallel 4-stream review (security, performance, architecture, testing) and a 5-phase comprehensive review pipeline
  • Session continuity — handoff documents with staleness classification for picking up exactly where you left off

On this page