The project
Technologies
Updated: March 2026
Tech stack
The site is a monolithic web application with server-side rendering and client-side hydration, built on these main technologies:
Backend
- Laravel — PHP framework
- Inertia.js — server/client bridge without a REST API
- MySQL — relational database
- Filament — admin panel
Frontend
- Vue 3 — reactive JavaScript framework
- TailwindCSS — utility-first CSS
- Leaflet.js — interactive maps
- Vite — bundler and dev server
Inertia.js SSR ensures fast load times and good search-engine indexability, while keeping a smooth client-side user experience.
AI tools
Some artificial intelligence tools were used in the process of building the site and writing the content.
Claude Code — software development
Anthropic AI agent used as a programming assistant: writing and reviewing code, component design, debugging, refactoring. The resulting code is always reviewed and validated by the author.
Anthropic LLM models — botanical texts
The descriptive species sheets were produced with the assistance of language models from printed scientific sources (Flora d'Italia, Flora analitica della Toscana, etc.) owned by the author. The final text is manually reviewed and verified. Each sheet explicitly states the presence of AI-assisted content.
The use of AI is a support tool, not a substitute for botanical knowledge. All floristic identifications and photographs are the author's.
Development methodology
The site is developed following the Spec-Driven Development (SDD) methodology with the OpenSpec tool: every new feature follows a formal cycle of proposal → specification → design → implementation → archiving.
- proposeThe desired change is described: why, what changes, which features it introduces
- designKey architectural decisions are made, alternatives and risks are assessed
- specsVerifiable requirements are written with WHEN/THEN scenarios
- applyIt is implemented task by task, marking progress
- archiveThe change is archived and the specs promoted to the main catalogue
This approach, applied also with the assistance of Claude Code, ensures that every change is motivated, documented and traceable over time.
Open data
The catalogue data is partially accessible in open format.
Sheets in Markdown format
Every species sheet and every botanical family is available in Markdown format by adding .md to the URL:
- /specie/{slug}.md
- /famiglia/{slug}.md
Index for LLMs
The /llms.txt file exposes an index of the catalogue readable by language models and AI agents, following the emerging llms.txt convention.
Public API — in development
A public API is under development that will allow structured access to the catalogue's botanical data (species, families, sightings, life forms). Details will be published here as soon as they are available.
The site content is released under the CC BY-NC-SA 4.0 licence: free for non-commercial use with attribution to piante-spontanee.it.
About this site · Accessibility · Privacy Policy · Updated: March 2026