Journal Archive — Page 161 of 166

SSE Makes Watchdog Necessary to Force Server Live-Reload Restore FastHTML/Uvicorn live-reload broken by SSE. Use Python watchdog to monitor files & force reliable server restarts for a smooth dev workflow.
API-Enabling the LLM Ghost Learn how an LLM can control browser DOM updates for CRUD actions via HTMX, WebSockets, and SSE, creating a seamless UI interaction mimicking user input.
How I Finally Got My LLM to Play Nice With the Web UI Debugging LLM-driven UI updates in FastHTML/HTMX. Switched from WebSockets to Server-Sent Events (SSE) for reliable backend-to-frontend command triggering.
FastHTML Hello World FastHTML 'Hello World' tutorial: Create web pages using Python functions (e.g., `H1()`, `Body()`) that map to HTML elements, skipping template languages.
Deep Debugging LLM Ghost Debugging LLM actions causing UI updates in FastHTML/HTMX. Chronicles WebSocket issues & proposes Server-Sent Events (SSE) as a robust solution.
Rabbit Hole Dev to Delight Clients Developer's reflection on balancing deep AI coding focus ('rabbit hole') with client needs. Strategies for resilience, avoiding distractions & managing...
Giving Gemini Advanced a Try Debugging LLM 'Ghost In The Machine' actions in FastHTML/HTMX. Covers AI-assisted refactoring, enhanced logging via decorators & NixOS helper scripts.
It's About Delighting Customers Prioritize customer delight (Drucker/Deming) in AI-assisted development. Refines workflow via Jupyter prototyping & structured client meeting prep.
LLM as Ambient App Intelligence Explore ambient LLM intelligence in apps: Details streaming WebSocket chat via Python/FastHTML & prioritizing client value (Drucker/Deming) in AI dev.
FastMTML Static Resources Manage static resources in FastHTML: Disable CDN defaults via `fast_app(default_hdrs=False)` and use `hdrs=(...)` for local JS/CSS includes.