Help

The command center. At a glance you can see total sites, active sites, vault entries, and any open alerts.

• Stat cards link directly to the relevant section — click Total Sites to go to the sites list.
• Quick actions surface the four most common tasks: add a site, open vault, configure Claude, and evaluate a URL.
• Client links let you share the plan wizard or site evaluator with a client directly.

The full list of client sites registered in the system. Each row shows the site name, URL, status, and client name.

• Click a site row to open its detail page.
• From the detail page you can view the blueprint canvas, trigger a new build, or edit site info.
• Use the "New Site" button (top right) to register a new client — only takes a minute.
• Status options: Active, Building, Maintenance, or Archived.

A visual canvas built on xyflow/react for mapping out a site's architecture. Each card represents a page, route, or component.

• Drag cards to reorganize the layout — positions save automatically to Supabase Storage.
• Use the canvas to plan the site before any code is written — pages, API routes, components, and their relationships.
• Blueprint data is referenced by the AI build pipeline when generating code.
• Access the blueprint from any site's detail page.

An AES-256-GCM encrypted credential store. Use it to store API keys, passwords, and environment variables per site.

• Nothing is stored in plaintext — all values are encrypted before hitting the database.
• Access requires a vault PIN that is never transmitted to the server.
• Credentials are scoped per site so each client's secrets stay separate.
• Add entries via Vault → Add, or directly from a site's detail page.

Trigger and monitor AI-assisted site builds. The build pipeline runs Claude through a series of phases to generate, refine, and ship code.

• Each build shows real-time progress, logs, and which phase is active.
• Iteration history is stored so you can compare before/after across build cycles.
• Builds pull from the site's blueprint and vault credentials automatically.
• Start a build from the Sites detail page or directly from the Builds section.

Run build or audit tasks across multiple sites at once. Useful when you've updated a shared pattern and want to apply it everywhere.

• Select the sites you want to target, choose the task type, then dispatch.
• Each site runs independently — one failure doesn't block others.
• Progress for each site is tracked separately in the build log.

SEO, security, and performance audits for any public URL. Results are stored per site for comparison over time.

• Audits run a suite of checks: title/meta, canonical, OG tags, robots, sitemap, HTTPS, redirect chains, and more.
• Each check is scored and flagged as pass, warn, or critical.
• The blueprint auto-generates fix nodes from audit findings so you can resolve issues in the build pipeline.
• Share the evaluate link at /evaluate with a client to let them kick off an audit themselves.

Active maintenance requests, mod installations, and the dispatch queue. Tracks ongoing work for sites that are in a support contract.

• Each request shows the site, description, status, and priority.
• Maintenance tasks can trigger targeted builds without a full rebuild.
• The dispatch queue integrates with Silver Creek Logistics for driver notification workflows.

Feedback analytics dashboard for the LinguaLens language learning app. Tracks learner XP, rank tiers, and feature feedback.

• View submitted feedback by learner tier (Beginner → Maestro) and rank tier (Bronze → Unreal).
• Use the feedback list to surface patterns across users before prioritizing features.
• Access the full feedback thread from any entry.

Visual iteration history for sites under active development. Each iteration captures a snapshot of what was built, scored, and changed.

• Compare iterations side-by-side to see visual and score progression.
• Iteration records include build notes, score deltas, and screenshot references.
• Useful for client check-ins — show exactly what changed between sessions.

Generate CLAUDE.md and Claude Code settings files tailored to a project. Also includes the workspace restore script and slash command reference.

• Choose a template (Next.js, WordPress, General) and fill in project details.
• Toggle Supabase, Tailwind, TypeScript, and Apple Aesthetic mode.
• Download both CLAUDE.md and settings.json as a ready-to-drop package.
• The Plugin & Skill Set tab gives you a one-command script to restore the full workspace on a new machine.
• The Slash Commands tab is a live reference for every /command across GStack, Superpowers, Vercel, and Ruflo.

Internal research tool for running structured analysis tasks. Used for competitive research, market sizing, and strategic synthesis.

• Input a topic or URL and select the analysis type.
• Results are structured and exportable for use in planning documents.
• Access is internal — not shared with clients.

Incoming submissions from client-facing forms, the plan wizard, and the evaluate page. Every lead or request lands here.

• Each submission shows the client name, contact info, site URL, and submitted data.
• Convert a submission into a new Site entry directly from the detail view.
• Submissions are never auto-deleted — archived entries stay searchable.

Mod installation tracking. Mods are optional add-ons that extend a site's functionality — analytics integrations, chat widgets, booking systems, and so on.

• Each mod shows installation status across all sites.
• Use the mod detail page to see which sites have it active, pending, or skipped.
• Mod installs can be triggered as part of a maintenance task or standalone build.