01
Automated generation with an AI chain
Queued jobs try providers in configurable order. On rate limit they fall back to the next; if all are exhausted, they defer and retry.
Back to projects
autobot.constructhub.org
Autobot
Web content generated and curated with AI
A platform of web-development guides and courses, with automated article generation via a multi-provider AI chain and an editorial calendar.
Technology
Stack used to build the product.
- PHP 8.2+ and Laravel 12
- Blade, Bootstrap 5, and Vite
- MySQL / MariaDB
- Redis/Valkey queues (database locally)
- AI chain: Gemini, Groq, Cerebras, OpenAI, Ollama
- Web-research fallback (Wikipedia, MDN, SearX/DuckDuckGo)
Architecture
How the system is structured: interfaces, backend, infrastructure, and operational flow.
Features
What the product does, in detail.
02
Automated editorial calendar
Topics (guides, lessons, spinoffs) scheduled on a calendar, with daily refill, continuous or batch dispatch, and weekly AI planning of new topics.
03
Courses, modules, and lessons
Course catalog with modules and lessons. Lessons enter the editorial queue in strict order and can auto-generate related spinoff articles.
04
Admin: review and publishing
Protected admin area for dashboard, articles (publish/reject/regenerate), categories, tags, courses, editorial plan, AI jobs, and API logs.
05
External content-generation API
Bearer endpoint that returns meta title, description, and HTML/Markdown synchronously or async (with polling and optional email), without publishing on the site.
06
SEO-ready public site
Home, articles, categories, tags, search, courses and lessons, sitemap.xml, SEO/Open Graph meta, and content attribution.
07
LLM-free web research fallback
When LLMs are unavailable, a web provider researches and synthesizes from trusted sources such as Wikipedia and MDN.
08
AI job monitoring and analytics
AI job and log tracking with stale-job retry, plus page-view and tile-click analytics viewable in admin.