Updated Readme.md for submission

docs(readme): add demo video, social post, and Substack links + credits

fix(oracle): on a parse failure, retry with the reliable baseline prompt

feat: in-app example moments + refined prompt on the Space

docs(readme): hackathon submission write-up, track + badge tags

revert(space): back to Qwen2.5-VL-7B after 3B failed verification

fix(oracle): retry when fewer than three films come back

feat(space): switch to Qwen2.5-VL-3B (small-model ethos, Tiny Titan)

fix(prompt): require the year field so it stops intermittently dropping

fix(prompt): strip year from title, harden JSON-only instruction

test: add live-Space A/B harness for baseline vs refined prompt

fix(backend): first pass respects refined env flag; add A/B override

feat(space): wire Qwen2.5-VL-7B via hf_local provider (stage 2)

build(space): adopt proven Gradio launch pattern for ZeroGPU

build(space): stage 1 — deployable shell for HF Space

feat(frontend): cinematic loading, page-load entrance, mobile polish

test: add prompt/model A/B comparison script

feat: avoid repeating films across a session

feat(backend): resilience — timeouts, bounded retry, clean errors

feat(backend): sanitize and re-encode all input before the model

feat(backend): retry once when an interpretation reads flat

feat(backend): make the typed fragment a primary signal

docs(backend): record proven ZeroGPU pattern, resolve spaces.GPU risk

test: add local smoke test for the oracle against LM Studio

feat(frontend): add Vite and React Chalchitra interface

feat(server): serve React UI and interpret API from gr.Server

feat(backend): add pluggable LLM oracle for image and fragment input

chore: add gitignore, env example, and requirements

initial commit