indigo-blog

As a frontend developer, I want to implement the Article detail page based on the existing Article (v2) JSX design so that guests and users can read a full published blog post. The page includes: article title, AI-generated hero image, full article body, author/date metadata, and a back-to-blog navigation link. Accessible from the Blog page by guests and from Articles page by users.

As a frontend developer, I want to implement the Login page based on the existing Login (v2) JSX design so that users and admins can sign in. The page includes email/password fields, validation, and navigation to the Dashboard on success. Follows the indigo-blog theme. Remove the scaffold login page and replace it with this design. Links from Landing page.

As a backend developer, I want to implement REST API endpoints for topic and source management so that users can create, read, update, and delete topics and their associated sources. Endpoints: POST /topics, GET /topics, DELETE /topics/{id}, POST /topics/{id}/sources, DELETE /topics/{id}/sources/{source_id}. Uses MySQL via SQLAlchemy and Alembic migrations.

As a frontend developer, I want to implement the Landing page based on the existing Landing (v2) JSX design so that guests, users, and admins see an interactive AI workspace homepage. The page includes: a dynamic topic map with 3D floating nodes, a glowing AI orb animation for live article generation, a live image creation showcase, micro-interactions on hover, and a publishing celebration confetti animation. Entry point for all personas — links to Login and Blog. Remove any pre-existing scaffold home page and replace with this design.

As a frontend developer, I want to implement the Blog page based on the existing Blog (v2) JSX design so that guests can browse and search published articles. The page includes: a searchable article grid/list, category/tag filters, article cards with AI-generated thumbnails, and links to individual Article detail pages. Also provides a link to Login for guests wishing to sign in.

As an AI engineer, I want to implement the end-to-end article generation pipeline so that the system can automatically search sources for given topics, generate article summaries, deduplicate via WeaviateDB, write full articles using GPT 5.2 via Litellm/Langchain, generate images using Google Nano Banana, and publish articles. Pipeline steps: 1) Source scraping per topic, 2) Story synthesis with GPT 5.2, 3) Summary generation + WeaviateDB deduplication check, 4) Full article generation if no duplicate, 5) Image generation, 6) Publish + save to WeaviateDB.

As a backend developer, I want to implement REST API endpoints for rate limit and interval configuration so that users can save and retrieve their generation preferences. Endpoints: GET /settings/rate-limits, PUT /settings/rate-limits, GET /settings/intervals, PUT /settings/intervals. Validates ranges per SRD constraints and stores in MySQL.

As a backend developer, I want to implement REST API endpoints for admin user management so that admins can list, view, update roles, suspend, and delete user accounts. Endpoints: GET /admin/users, GET /admin/users/{id}, PATCH /admin/users/{id}, POST /admin/users/{id}/suspend, DELETE /admin/users/{id}. Restricted to Admin role via RBAC middleware.

As a frontend developer, I want to implement the Dashboard page based on the existing Dashboard (v2) JSX design so that users and admins can view an overview of their stats, recent articles, and system activity. The page includes: stat cards, recent articles list, generation status indicators, and navigation to Topics, Articles, and Settings. Admin view also shows system monitoring and logs. Entry point after login.

As a frontend developer, I want to implement the Users page based on the existing Users (v2) JSX design so that admins can manage user accounts. The page includes: a searchable user table with role badges (Admin/User/Guest), status indicators, action menus (edit, suspend, delete), a user detail slide-in drawer with profile, activity, and permissions tabs, and pagination. Links from Dashboard (admin flow only).

As a frontend developer, I want to implement the Topics page based on the existing Topics (v2) JSX design so that users can add, view, and manage their list of topics and associated sources. The page includes: a topic list panel, an 'Add Topic' form, a source management sub-panel per topic (add/remove sources), and status indicators for each topic. Links from Dashboard; leads to Settings for rate limit configuration.

As a backend developer, I want to implement REST API endpoints for article management so that users can list, retrieve, edit, and publish generated articles. Endpoints: GET /articles, GET /articles/{id}, PATCH /articles/{id}, POST /articles/{id}/publish, DELETE /articles/{id}. Returns article metadata, content, status (draft/published/rejected), and AI-generated image URL.

As a frontend developer, I want to implement the Settings page based on the existing Settings (v2) JSX design so that users can configure rate limits and generation intervals. The page includes: a rate limit slider/input (max articles per day/week), interval frequency selector (hourly, daily, weekly, custom cron), active hours picker, and a save button. Admin view includes additional system configuration options. Links from Topics and Dashboard.

As a frontend developer, I want to implement the Articles page based on the existing Articles (v2) JSX design so that users can review AI-generated articles, edit them, and publish them. The page includes: a list/grid of generated articles with status badges (draft, published, duplicate-rejected), a preview panel, edit capability, and a publish action button. Admin can also audit all content. Links from Dashboard.

As a backend developer, I want to implement a scheduler service so that the system automatically triggers topic searches and article generation at user-configured intervals. Supports: hourly, every 6 hours, daily, weekly, and custom cron expressions. Respects per-user rate limits. Uses Celery or APScheduler with Redis/MySQL for job state. Exposes next-run timestamp via API.

As a data engineer, I want to integrate WeaviateDB into the article pipeline so that generated article summaries are checked for semantic duplicates before full generation, and published articles are saved for future deduplication. Includes: WeaviateDB schema setup for article vectors, similarity search query on summary embedding, insert-on-publish flow, and a deduplication threshold configuration.

As a backend developer, I want to implement REST API endpoints for dashboard statistics and system logs so that users see their generation overview and admins can monitor system activity. Endpoints: GET /dashboard/stats (articles generated, published, rate limit usage), GET /admin/logs (paginated system activity log). Aggregates data from MySQL and generation pipeline.

As a backend developer, I want to implement public REST API endpoints so that guests can browse, search, and read published blog articles without authentication. Endpoints: GET /blog/articles (with search and filter params), GET /blog/articles/{slug}. Returns published articles with full content, images, and metadata. No auth required.

As a frontend developer, I want to connect all implemented pages to their respective backend API endpoints so that the application is fully functional end-to-end. This includes: setting up an Axios/Fetch API client with auth token injection, connecting Topics page to Topics & Sources API, Articles page to Articles CRUD API, Settings page to Rate Limits & Intervals API, Dashboard to Stats API, Users page to User Management API, and Blog/Article pages to Public Blog API. Handle loading, error, and empty states on all pages.