How rudra00434/opencluely-v2.0 works
OpenCluely is an Electron interview copilot: users ask by voice or screenshot, Gemini reasons with skill prompts and session context, and answers stream to overlay or chat. The graph follows the documented workflow and sampled imports/calls; several renderer and integration details are unsampled, so unsupported connections are omitted.
Main entry points: main.js App Controller · llm.service.js Gemini Reasoning · speech.service.js Speech Capture · capture.service.js Screenshot Capture
Components 21
Desktop Control
- App Controller
main.js - Configuration
src/core/config.js - First Run
src/core/first-run.js - Whisper Setup
src/core/whisper-installer.js
Capture and Speech
- Screenshot Capture
src/services/capture.service.js - Speech Capture
src/services/speech.service.js - Whisper Worker
src/services/whisper-worker.service.js
AI and Session
- Session Memory
src/managers/session.manager.js - Skill Prompts
prompt-loader.js - Gemini Reasoning
src/services/llm.service.js
Windows and Support
- Window Manager
src/managers/window.manager.js - Main Window UI
src/ui/main-window.js - Chat Window UI
src/ui/chat-window.js - Settings Window UI
src/ui/settings-window.js - Answer Renderer
llm-response.html - Markdown Rendering
lib/markdown.js - Math Rendering
lib/mathrender.js - Logging
src/core/logger.js
Other components
- Interview User
- Azure Speech
- Gemini API
Connections 25
- Interview User → App Controller: asks questions
README.md - App Controller → Screenshot Capture: requests capture
main.js - App Controller → Speech Capture: controls capture
main.js - App Controller → Gemini Reasoning: submits questions
README.md - Screenshot Capture → Gemini Reasoning: sends image
README.md - Speech Capture → Whisper Worker: transcribes locally
src/services/speech.service.js - Speech Capture → Azure Speech: uses provider
README.md - Gemini Reasoning → Gemini API: requests answers
README.md - Gemini Reasoning → Session Memory: uses context
src/services/llm.service.js - Gemini Reasoning → Skill Prompts: loads prompts
src/services/llm.service.js - Session Memory → Skill Prompts: loads skill prompts
src/managers/session.manager.js - App Controller → Session Memory: manages session
main.js - App Controller → Skill Prompts: selects skills
main.js - App Controller → Window Manager: initializes windows
main.js - App Controller → First Run: checks onboarding
main.js - App Controller → Whisper Setup: sets up Whisper
main.js - Window Manager → Answer Renderer: loads response window
src/managers/window.manager.js - Answer Renderer → Markdown Rendering: renders markdown
- Answer Renderer → Math Rendering: renders math
- App Controller → Main Window UI: serves interface
- App Controller → Chat Window UI: routes voice answers
README.md - App Controller → Settings Window UI: configures preferences
- Speech Capture → Logging: logs events
src/services/speech.service.js - Screenshot Capture → Logging: logs capture
src/services/capture.service.js - App Controller → Configuration: reads settings
main.js