🤖 Developed & Engineered with AI: This application is fully designed, architected, and built with Advanced AI Pair Programming.
ClipFarm is a high-performance desktop application that automatically indexes your Medal.tv gameplay library across SQLite databases and custom storage folders, extracts in-game kill events (Kills, Headshots, Multi-kills, Aces), and cuts clean highlight reels with zero quality loss in under 0.1s.
- 🔍 Full Medal Library & SQLite Synchronization: Seamlessly scans both
%APPDATA%\Medal\store\clips.jsonand Medal's internal SQLite database (medal-*.db), discovering thousands of clips and exact ICYMI kill timestamps. - 📁 Multi-Source Folder Scanning: Add custom source directories (e.g.
E:\Videos\Medal\Clips,E:\Videos\Medal\Edits) to index your entire gameplay archive. - ⚡ Zero-Loss Stream Copy (0.1s Trims): Cuts clips instantaneously using direct stream copying without re-encoding, preserving 100% of the original video and audio quality.
- 🎮 Hardware GPU Acceleration: Detects AMD AMF (Radeon), NVIDIA NVENC (GeForce/RTX), and Intel QSV (Arc) for ultra-fast re-encoding when needed.
- 🎯 4 Versatile Trimming Modes:
- Clip Highlights (1 File Per Clip): Extracts all frags in each recording into a single consolidated highlight video per clip.
- Smart Merge (Recommended): Combines multi-kills occurring within seconds of each other into smooth highlight reels.
- Individual Files: Exports each frag as an isolated
.mp4clip. - Montage / Compilation: Stitches all selected clips into a single compilation video.
- 📂 Automatic Game-Subfolder Sorting: Automatically categorizes exported highlights into clean game folders (e.g.
Valorant/,Counter-Strike 2/). - 🧹 Noise-Free Telemetry Filtering: Filters out non-kill telemetry noise (
RoundPhaseChange,Bomb,Death,Assist), retaining exclusively real frags and highlight moments. - 🗑️ Automatic Source Cleanup (Optional): Option to automatically delete the raw Medal recording once the highlight is saved to free up disk space.
- 🚀 144 FPS Virtual Infinite Scroll: Renders thousands of clips smoothly without DOM lag using virtual pagination and chunked loading.
- 🎬 Interactive Timeline Player: Frame-accurate seeking, color-coded kill pins, and 1-click launch in your system media player.
- 🔄 Integrated GitHub Auto-Updater: Check for updates and view new releases directly inside the application settings.
- 🌐 Bilingual Interface: 1-click switch between English and Polish.
- Download
ClipFarm.exefrom Releases or compile it locally viabuild_exe.bat. - Launch
ClipFarm.exe— no extra setup or installation required!
- Prerequisites: Python 3.10+ and Node.js 18+ (optional, production dist is included)
- Launch with 1-click:
run_clipfarm.bat
- Manual terminal launch:
pip install -r requirements.txt python main.py
All settings can be customized through the in-app Settings modal and are persisted in config.json:
| Setting | Description | Default |
|---|---|---|
| Lead Time | Seconds of gameplay captured before the kill event | 3.0s |
| Tail Time | Seconds of gameplay captured after the kill event | 2.0s |
| Merge Gap | Maximum gap threshold to combine consecutive kills into one clip | 3.0s |
| Encoding Mode | stream_copy (lossless, 0s), amf (AMD), nvenc (NVIDIA), cpu (x264) |
stream_copy |
| Auto-Delete Source | Automatically delete raw source recording after trimming | Disabled |
| Output Directory | Destination directory for exported highlights | E:\Wideo\Medal\ClipFarm_Exports |
graph TD
A["Medal SQLite DB & clips.json"] --> B["ClipFarm Python Engine (Flask)"]
S["Custom Source Folders"] --> B
B --> C["Event Classifier (Kill, Headshot, Ace)"]
C --> D["Interval Optimizer & Telemetry Filter"]
D --> E["FFmpeg Engine (Stream Copy / AMF / NVENC)"]
E --> F["Organized Exports (Game Subfolders: Valorant/, CS2/)"]
B <--> G["Modern React 18 UI (Vite + TailwindCSS + i18n)"]
This project is open-source and licensed under the MIT License.