# sahilsleem/autocut architecture

> Architecture diagram of the GitHub repository sahilsleem/autocut, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-06). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/sahilsleem/autocut
- Repository: https://github.com/sahilsleem/autocut

## Overview

AutoCut is a unified video editor with separate long-form and vertical workflows. Users import footage, build or refine timelines, preview edits, and render finished video; the long-form workflow adds transcription, media analysis, semantic matching, and automated timeline drafting. The graph includes the documented worker/native processing options and persistent project state. Source excerpts and the index establish several client-side calls; some worker internals are unsampled, so their links are limited to documented or explicitly shown relationships.

Main entry points: [`src/state/useProjectStore.ts`](https://github.com/sahilsleem/autocut/blob/main/src/state/useProjectStore.ts) (Project state), [`android/app/src/main/java/com/longformai/app/MainActivity.java`](https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/java/com/longformai/app/MainActivity.java) (Native plugin registration), [`src/LongFormEditor.tsx`](https://github.com/sahilsleem/autocut/blob/main/src/LongFormEditor.tsx) (Long-form editor), [`src/App.tsx`](https://github.com/sahilsleem/autocut/blob/main/src/App.tsx) (Mode selection)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_editing["Editing workflows"]
  node_app["Mode selection<br/>[App.tsx]"]
  node_longform["Long-form editor<br/>[LongFormEditor.tsx]"]
  node_shorts["Vertical editor<br/>[ShortsEditor.tsx]"]
  node_timeline["Timeline and preview<br/>[Timeline.tsx]"]
  node_captions["Caption controls"]
end

subgraph group_intelligence["Media intelligence"]
  node_transcription["Speech transcription<br/>[transcription.ts]"]
  node_media_analysis["Media analysis<br/>[mediaAnalysis.ts]"]
  node_vision["Visual understanding<br/>[vision.ts]"]
  node_matching["Semantic matching<br/>[matching.ts]"]
  node_draft["Timeline drafting<br/>[draftTimeline.ts]"]
  node_preparation["Project preparation<br/>[preparation.ts]"]
  node_vision_worker["Vision worker<br/>[vision_server.py]"]
  node_matching_worker["Matching worker<br/>[matching_server.py]"]
  node_transcription_worker["Transcription worker"]
end

subgraph group_processing["Processing and render"]
  node_render_ui["Render controls<br/>[RenderModal.tsx]"]
  node_render["Render pipeline<br/>[render.ts]"]
  node_platform_render["Platform rendering<br/>[render.ts]"]
  node_render_worker["Render worker<br/>[render_server.py]"]
end

subgraph group_platform["Platform and persistence"]
  node_project_state["Project state<br/>[useProjectStore.ts]"]
  node_project_schema["Project and timeline schema<br/>[project.ts]"]
  node_persistence[("Project persistence<br/>[persistence.ts]")]
  node_worker_config["Worker routing<br/>[workerConfig.ts]"]
  node_worker_health["Worker diagnostics<br/>[workers.ts]"]
  node_android_media["Android media access<br/>[androidMedia.ts]"]
end

subgraph group_native["Android native engines"]
  node_main_activity["Native plugin registration<br/>[MainActivity.java]"]
  node_native_blip["Native BLIP engine"]
  node_tokenizer["WordPiece tokenizer"]
  node_native_whisper["Native Whisper plugin"]
  node_whisper_core["Whisper inference core<br/>[whisper.cpp]"]
  node_ggml["GGML tensor runtime<br/>[ggml.h]"]
  node_native_ffmpeg["Native FFmpeg plugin"]
  node_native_minilm["Native MiniLM plugin"]
end

node_creator(("Creator"))
node_finished_video["Finished video"]

node_creator -->|"chooses workflow"| node_app
node_app -->|"opens"| node_longform
node_app -->|"opens"| node_shorts
node_longform -->|"updates"| node_project_state
node_longform -->|"uses"| node_timeline
node_longform -->|"opens"| node_render_ui
node_shorts -->|"renders through"| node_platform_render
node_project_state -->|"transcribes audio"| node_transcription
node_project_state -->|"analyzes media"| node_media_analysis
node_project_state -->|"prepares project"| node_preparation
node_project_state -->|"saves project"| node_persistence
node_project_state -->|"holds"| node_project_schema
node_media_analysis -->|"requests semantics"| node_vision
node_draft -->|"ranks media"| node_matching
node_draft -->|"uses types"| node_project_schema
node_vision -->|"uses on Android"| node_native_blip
node_vision -.->|"requests analysis"| node_vision_worker
node_matching -->|"uses on Android"| node_native_minilm
node_matching -.->|"requests matches"| node_matching_worker
node_transcription -.->|"requests transcription"| node_transcription_worker
node_worker_health -->|"resolves endpoints"| node_worker_config
node_worker_health -->|"checks platform"| node_android_media
node_render_ui -.->|"starts rendering"| node_render
node_render -.->|"requests render"| node_render_worker
node_render -->|"produces"| node_finished_video
node_main_activity -->|"registers plugin"| node_native_blip
node_main_activity -->|"registers plugin"| node_native_whisper
node_main_activity -->|"registers plugin"| node_native_ffmpeg
node_main_activity -->|"registers plugin"| node_native_minilm
node_native_blip -->|"tokenizes text"| node_tokenizer
node_native_whisper -->|"bridges inference"| node_whisper_core
node_whisper_core -->|"uses tensor runtime"| node_ggml

click node_app "https://github.com/sahilsleem/autocut/blob/main/src/App.tsx"
click node_longform "https://github.com/sahilsleem/autocut/blob/main/src/LongFormEditor.tsx"
click node_shorts "https://github.com/sahilsleem/autocut/blob/main/src/ShortsEditor.tsx"
click node_timeline "https://github.com/sahilsleem/autocut/blob/main/src/components/Timeline.tsx"
click node_captions "https://github.com/sahilsleem/autocut/blob/main/src/components/TranscriptPanel.tsx"
click node_project_state "https://github.com/sahilsleem/autocut/blob/main/src/state/useProjectStore.ts"
click node_project_schema "https://github.com/sahilsleem/autocut/blob/main/src/types/project.ts"
click node_persistence "https://github.com/sahilsleem/autocut/blob/main/src/engine/persistence.ts"
click node_transcription "https://github.com/sahilsleem/autocut/blob/main/src/engine/transcription.ts"
click node_media_analysis "https://github.com/sahilsleem/autocut/blob/main/src/engine/mediaAnalysis.ts"
click node_vision "https://github.com/sahilsleem/autocut/blob/main/src/engine/vision.ts"
click node_matching "https://github.com/sahilsleem/autocut/blob/main/src/engine/matching.ts"
click node_draft "https://github.com/sahilsleem/autocut/blob/main/src/engine/draftTimeline.ts"
click node_preparation "https://github.com/sahilsleem/autocut/blob/main/src/engine/preparation.ts"
click node_render_ui "https://github.com/sahilsleem/autocut/blob/main/src/components/RenderModal.tsx"
click node_render "https://github.com/sahilsleem/autocut/blob/main/src/engine/render.ts"
click node_platform_render "https://github.com/sahilsleem/autocut/blob/main/src/platform/render.ts"
click node_worker_config "https://github.com/sahilsleem/autocut/blob/main/src/config/workerConfig.ts"
click node_worker_health "https://github.com/sahilsleem/autocut/blob/main/src/engine/workers.ts"
click node_android_media "https://github.com/sahilsleem/autocut/blob/main/src/platform/androidMedia.ts"
click node_vision_worker "https://github.com/sahilsleem/autocut/blob/main/server/vision_server.py"
click node_matching_worker "https://github.com/sahilsleem/autocut/blob/main/server/matching_server.py"
click node_transcription_worker "https://github.com/sahilsleem/autocut/blob/main/server/transcribe_server.py"
click node_render_worker "https://github.com/sahilsleem/autocut/blob/main/server/render_server.py"
click node_main_activity "https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/java/com/longformai/app/MainActivity.java"
click node_native_blip "https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/java/com/longformai/app/NativeBlipEngine.java"
click node_tokenizer "https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/java/com/longformai/app/BertWordPieceTokenizer.java"
click node_native_whisper "https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/java/com/longformai/app/NativeWhisperPlugin.java"
click node_whisper_core "https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/cpp/whisper.cpp"
click node_ggml "https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/cpp/ggml.h"
click node_native_ffmpeg "https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/java/com/longformai/app/NativeFFmpegPlugin.java"
click node_native_minilm "https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/java/com/longformai/app/NativeMiniLMPlugin.java"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_app,node_longform,node_shorts,node_timeline,node_captions toneBlue
class node_transcription,node_media_analysis,node_vision,node_matching,node_draft,node_preparation,node_vision_worker,node_matching_worker,node_transcription_worker toneAmber
class node_render_ui,node_render,node_platform_render,node_render_worker toneMint
class node_project_state,node_project_schema,node_persistence,node_worker_config,node_worker_health,node_android_media toneRose
class node_main_activity,node_native_blip,node_tokenizer,node_native_whisper,node_whisper_core,node_ggml,node_native_ffmpeg,node_native_minilm,node_creator,node_finished_video toneIndigo
```

## Components

### Editing workflows

- [Mode selection](https://github.com/sahilsleem/autocut/blob/main/src/App.tsx): `src/App.tsx`
- [Long-form editor](https://github.com/sahilsleem/autocut/blob/main/src/LongFormEditor.tsx): `src/LongFormEditor.tsx`
- [Vertical editor](https://github.com/sahilsleem/autocut/blob/main/src/ShortsEditor.tsx): `src/ShortsEditor.tsx`
- [Timeline and preview](https://github.com/sahilsleem/autocut/blob/main/src/components/Timeline.tsx): `src/components/Timeline.tsx`
- [Caption controls](https://github.com/sahilsleem/autocut/blob/main/src/components/TranscriptPanel.tsx): `src/components/TranscriptPanel.tsx`

### Media intelligence

- [Speech transcription](https://github.com/sahilsleem/autocut/blob/main/src/engine/transcription.ts): `src/engine/transcription.ts`
- [Media analysis](https://github.com/sahilsleem/autocut/blob/main/src/engine/mediaAnalysis.ts): `src/engine/mediaAnalysis.ts`
- [Visual understanding](https://github.com/sahilsleem/autocut/blob/main/src/engine/vision.ts): `src/engine/vision.ts`
- [Semantic matching](https://github.com/sahilsleem/autocut/blob/main/src/engine/matching.ts): `src/engine/matching.ts`
- [Timeline drafting](https://github.com/sahilsleem/autocut/blob/main/src/engine/draftTimeline.ts): `src/engine/draftTimeline.ts`
- [Project preparation](https://github.com/sahilsleem/autocut/blob/main/src/engine/preparation.ts): `src/engine/preparation.ts`
- [Vision worker](https://github.com/sahilsleem/autocut/blob/main/server/vision_server.py): `server/vision_server.py`
- [Matching worker](https://github.com/sahilsleem/autocut/blob/main/server/matching_server.py): `server/matching_server.py`
- [Transcription worker](https://github.com/sahilsleem/autocut/blob/main/server/transcribe_server.py): `server/transcribe_server.py`

### Processing and render

- [Render controls](https://github.com/sahilsleem/autocut/blob/main/src/components/RenderModal.tsx): `src/components/RenderModal.tsx`
- [Render pipeline](https://github.com/sahilsleem/autocut/blob/main/src/engine/render.ts): `src/engine/render.ts`
- [Platform rendering](https://github.com/sahilsleem/autocut/blob/main/src/platform/render.ts): `src/platform/render.ts`
- [Render worker](https://github.com/sahilsleem/autocut/blob/main/server/render_server.py): `server/render_server.py`

### Platform and persistence

- [Project state](https://github.com/sahilsleem/autocut/blob/main/src/state/useProjectStore.ts): `src/state/useProjectStore.ts`
- [Project and timeline schema](https://github.com/sahilsleem/autocut/blob/main/src/types/project.ts): `src/types/project.ts`
- [Project persistence](https://github.com/sahilsleem/autocut/blob/main/src/engine/persistence.ts): `src/engine/persistence.ts`
- [Worker routing](https://github.com/sahilsleem/autocut/blob/main/src/config/workerConfig.ts): `src/config/workerConfig.ts`
- [Worker diagnostics](https://github.com/sahilsleem/autocut/blob/main/src/engine/workers.ts): `src/engine/workers.ts`
- [Android media access](https://github.com/sahilsleem/autocut/blob/main/src/platform/androidMedia.ts): `src/platform/androidMedia.ts`

### Android native engines

- [Native plugin registration](https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/java/com/longformai/app/MainActivity.java): `android/app/src/main/java/com/longformai/app/MainActivity.java`
- [Native BLIP engine](https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/java/com/longformai/app/NativeBlipEngine.java): `android/app/src/main/java/com/longformai/app/NativeBlipEngine.java`
- [WordPiece tokenizer](https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/java/com/longformai/app/BertWordPieceTokenizer.java): `android/app/src/main/java/com/longformai/app/BertWordPieceTokenizer.java`
- [Native Whisper plugin](https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/java/com/longformai/app/NativeWhisperPlugin.java): `android/app/src/main/java/com/longformai/app/NativeWhisperPlugin.java`
- [Whisper inference core](https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/cpp/whisper.cpp): `android/app/src/main/cpp/whisper.cpp`
- [GGML tensor runtime](https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/cpp/ggml.h): `android/app/src/main/cpp/ggml.h`
- [Native FFmpeg plugin](https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/java/com/longformai/app/NativeFFmpegPlugin.java): `android/app/src/main/java/com/longformai/app/NativeFFmpegPlugin.java`
- [Native MiniLM plugin](https://github.com/sahilsleem/autocut/blob/main/android/app/src/main/java/com/longformai/app/NativeMiniLMPlugin.java): `android/app/src/main/java/com/longformai/app/NativeMiniLMPlugin.java`

### Other components

- **Creator**
- **Finished video**

## Connections

- Creator → Mode selection: chooses workflow (evidence: `src/App.tsx`)
- Mode selection → Long-form editor: opens (evidence: `src/App.tsx`)
- Mode selection → Vertical editor: opens (evidence: `src/App.tsx`)
- Long-form editor → Project state: updates (evidence: `src/LongFormEditor.tsx`)
- Long-form editor → Timeline and preview: uses (evidence: `src/LongFormEditor.tsx`)
- Long-form editor → Render controls: opens (evidence: `src/LongFormEditor.tsx`)
- Vertical editor → Platform rendering: renders through (evidence: `src/ShortsEditor.tsx`)
- Project state → Speech transcription: transcribes audio (evidence: `src/state/useProjectStore.ts`)
- Project state → Media analysis: analyzes media (evidence: `src/state/useProjectStore.ts`)
- Project state → Project preparation: prepares project (evidence: `src/state/useProjectStore.ts`)
- Project state → Project persistence: saves project (evidence: `src/state/useProjectStore.ts`)
- Project state → Project and timeline schema: holds (evidence: `src/state/useProjectStore.ts`)
- Media analysis → Visual understanding: requests semantics (evidence: `src/engine/mediaAnalysis.ts`)
- Timeline drafting → Semantic matching: ranks media (evidence: `src/engine/draftTimeline.ts`)
- Timeline drafting → Project and timeline schema: uses types (evidence: `src/engine/draftTimeline.ts`)
- Visual understanding → Native BLIP engine: uses on Android (evidence: `src/engine/vision.ts`)
- Visual understanding → Vision worker: requests analysis
- Semantic matching → Native MiniLM plugin: uses on Android (evidence: `src/engine/matching.ts`)
- Semantic matching → Matching worker: requests matches
- Speech transcription → Transcription worker: requests transcription
- Worker diagnostics → Worker routing: resolves endpoints (evidence: `src/engine/workers.ts`)
- Worker diagnostics → Android media access: checks platform (evidence: `src/engine/workers.ts`)
- Render controls → Render pipeline: starts rendering
- Render pipeline → Render worker: requests render
- Render pipeline → Finished video: produces
- Native plugin registration → Native BLIP engine: registers plugin (evidence: `android/app/src/main/java/com/longformai/app/MainActivity.java`)
- Native plugin registration → Native Whisper plugin: registers plugin (evidence: `android/app/src/main/java/com/longformai/app/MainActivity.java`)
- Native plugin registration → Native FFmpeg plugin: registers plugin (evidence: `android/app/src/main/java/com/longformai/app/MainActivity.java`)
- Native plugin registration → Native MiniLM plugin: registers plugin (evidence: `android/app/src/main/java/com/longformai/app/MainActivity.java`)
- Native BLIP engine → WordPiece tokenizer: tokenizes text (evidence: `android/app/src/main/java/com/longformai/app/NativeBlipEngine.java`)
- Native Whisper plugin → Whisper inference core: bridges inference (evidence: `android/app/src/main/java/com/longformai/app/NativeWhisperPlugin.java`)
- Whisper inference core → GGML tensor runtime: uses tensor runtime (evidence: `android/app/src/main/cpp/whisper.cpp`)
