## 1. Repo And Tooling Setup

- [x] 1.1 Choose the site framework and initialize the project skeleton (Astro preferred) with TypeScript support
- [x] 1.2 Add a consistent config approach for secrets and runtime settings (env vars for API keys, Umami settings, base URL)
- [x] 1.3 Add basic quality gates (lint/format, typecheck) and a minimal CI workflow
- [x] 1.4 Add Docker build/run for deployment to the Linode instance (multi-stage build, static output served by nginx or equivalent)

## 2. Content Model And Cache

- [x] 2.1 Define the normalized content item schema (id, source, url, title, publishedAt, thumbnailUrl, optional metrics/views)
- [x] 2.2 Implement a content cache file format and location (e.g., `content/cache/*.json`) plus a "last updated" timestamp
- [x] 2.3 Implement cache read utilities used by page rendering, including source filtering and limit controls
- [x] 2.4 Implement "serve last good cache on failure" behavior in the ingestion workflow (write temp then swap)

## 3. YouTube Ingestion

- [x] 3.1 Implement YouTube ingestion via RSS as the baseline (latest videos, no metrics)
- [x] 3.2 Implement optional YouTube Data API ingestion when `YOUTUBE_API_KEY` is configured (include `metrics.views`)
- [x] 3.3 Implement configurable channel identifier for `youtube.com/santhoshj` (document required values)
- [x] 3.4 Add a small set of local fixtures/tests for parsing and normalization (RSS and API response samples)

## 4. Podcast Ingestion

- [x] 4.1 Add configuration for the Irregular Mind podcast RSS URL
- [x] 4.2 Implement podcast RSS fetch + parse to normalized items (one item per episode)
- [x] 4.3 Ensure episode URLs and publish dates normalize correctly across common RSS date formats
- [x] 4.4 Add fixtures/tests for RSS parsing and normalization

## 5. Instagram Support (Embed-First)

- [x] 5.1 Define the minimal normalized representation for Instagram items in embed-first mode (post URL + optional thumbnail)
- [x] 5.2 Implement a configuration-driven list of Instagram post URLs (or a simple feed source) for v1
- [x] 5.3 Implement the renderer for Instagram embeds with a graceful empty state when no items exist
- [x] 5.4 Document a Phase 2 path for API-based ingestion (constraints, account/app requirements, token refresh)

## 6. Build-Time Refresh And Scheduling

- [x] 6.1 Implement a `fetch-content` script/command that ingests all sources and writes the cache
- [x] 6.2 Add a manual refresh mechanism for ops (document how to run on the Linode host)
- [x] 6.3 Add a scheduled refresh plan (cron on Linode or CI scheduler) and document frequency (daily minimum)
- [x] 6.4 Add logging for ingestion runs (success/failure per source, cache write result)

## 7. Pages And SEO Surface

- [x] 7.1 Implement `/` home page scaffold with deterministic module order
- [x] 7.2 Implement `/videos`, `/podcast`, and `/about` indexable pages (static/SSR HTML, not client-only)
- [x] 7.3 Add per-page metadata: title, description, canonical URL
- [x] 7.4 Add Open Graph + Twitter card tags for indexable pages
- [x] 7.5 Implement `sitemap.xml` and `robots.txt` (include `/`, `/videos`, `/podcast`, `/about`)
- [x] 7.6 Add structured data (JSON-LD) scaffolding for Video/Podcast where detail pages exist (ensure valid JSON)

## 8. Homepage Modules

- [x] 8.1 Implement "newest content" feed (sort by `publishedAt` desc across sources, configurable limits)
- [x] 8.2 Implement "high-performing videos" module using `metrics.views` when present
- [x] 8.3 Implement a manual override/curation list for featured/high-performing videos
- [x] 8.4 Implement graceful empty/error states for each module (page never breaks if a module has no data)

## 9. CTAs And Outbound Tracking

- [x] 9.1 Implement a reusable CTA component with `platform`, `placement`, and destination `url`
- [x] 9.2 Add default destinations for YouTube channel, Instagram profile, and podcast listen target(s)
- [x] 9.3 Add UTM parameter support for CTA outbound links
- [x] 9.4 Define CTA placements on homepage (hero + at least one additional placement) and ensure they are visually prominent

## 10. Umami Analytics

- [x] 10.1 Add Umami configuration (website ID + script URL) via environment/config without code changes per environment
- [x] 10.2 Load Umami on all indexable pages when enabled; ensure site works when disabled
- [x] 10.3 Implement `cta_click` event emission with `platform`, `placement`, and `target`
- [x] 10.4 Implement `outbound_click` event emission with destination domain/identifier

## 11. Verification And Launch

- [x] 11.1 Validate SEO basics locally (view-source contains metadata; sitemap/robots served correctly)
- [x] 11.2 Validate performance basics (no blocking API calls at runtime; pages render from cache)
- [x] 11.3 Validate analytics in a staging environment (events show up in Umami; disabled mode has no errors)
- [x] 11.4 Document deployment steps for the Linode Docker setup (build, run, refresh schedule, rollback)