Project Overview

Welcome to the Llama Chile Shop (LCS) Pipeline — an automated system for rendering, titling, uploading, and archiving Fortnite gameplay highlights starring Gramps!

✅ Current Status

The system is fully operational for montage and single-clip rendering:

Auto-discovers clips from montages/, hits/, misses/, outtakes/
Renders each clip with intro, dynamic title card, and outro
Uploads to YouTube with appropriate playlist routing
Archives structured metadata, including URLs and tags
Supports dynamic description generation via OpenAI

📊 Feature Snapshot

Feature	Status	Notes
Clip rendering (wide/vert)	✅ Done	Fully automated
YouTube upload	✅ Done	Includes playlist routing, category ID, hashtags
Dynamic metadata generation	✅ Done	OpenAI template logic working
Metadata archive	✅ Done	Saved locally with YouTube/PeerTube URLs
Mastodon posting	🧪 In progress	Posting logic scaffolded, needs integration
PeerTube upload	⏳ Planned	Awaiting config/token/queueing
`outtakes/` manual support	✅ Implemented	Requires manual trigger
Wiki rendering	⚠️ Blocked	GitHub refuses to publish updates via API
`sync_wiki.py`	🪦 Deprecated	Kept as emergency parachute
Metadata model v2	✅ Done	Supports wide + vertical, multiple clips per session

🗂️ Folder Structure

Each stream session lives under Z:\YYYY.MM.DD or Z:\YYYY.MM.DD.N, with these subfolders:

hits/ — short action highlight clips
misses/ — funny fails
montages/ — pre-trimmed reels
outtakes/ — manually marked events
rendered/ — auto-generated videos go here

🧠 Current Version

Version: v0.1.1 (YouTube + render + metadata complete)
Next milestone: v0.1.2 (PeerTube support)
Source: C:\Users\alpha\code\Python\lcs_agent_clean
Mount: NAS path is Z:\ (mapped to \\chong\LCS\Videos\eklipse)

🔧 Development Tips

To test safely, set DEBUG = True in config.py
Title overlays use Fortnite fonts/colors
notes.json → OpenAI → Upload → Archive → Social post

🧠 Maintained by ChatGPT, your resident senior dev + wiki wrangler. 😎