Intent-driven AI video content generation pipeline with approval-gated workflow.
- Rust 74.9%
- HTML 19%
- CSS 5.4%
- Makefile 0.7%
|
|
||
|---|---|---|
| .forgejo/workflows | ||
| crates | ||
| docs | ||
| .gitignore | ||
| Cargo.toml | ||
| Cargo.toml.hero_builder_backup | ||
| LICENSE | ||
| Makefile | ||
| PURPOSE.md | ||
| README.md | ||
| rust-toolchain.toml | ||
hero_videos
AI video content generation pipeline — intent-driven, approval-gated workflow from idea to assembled video.
See PURPOSE.md for component overview.
Quick Start
lab build hero_videos --install --start
Service Lifecycle
# Build and install all binaries
lab build hero_videos --install
# Start via hero_proc
lab build hero_videos --start
# Check status
hero_proc list
Pipeline
- Scene Planning — describe your intent, AI generates scene prompts
- Image Generation — OpenRouter/Gemini generates candidate images per scene
- Video Generation — OpenRouter/Kling animates each selected image
- Assembly — ffmpeg concatenates approved clips into a final downloadable video
Required Secrets
lab secrets set OPENROUTER_API_KEY <key> # image + video generation
lab secrets set GROQ_API_KEY <key> # transcription (optional)
Documentation
- PURPOSE.md — component overview and provider table
- docs/ — OpenRPC spec and schema reference