pluely

Abstract

The Open Source Alternative to Cluely - A lightning-fast, privacy-first AI assistant that works seamlessly during meetings, interviews, and conversations without anyone knowing. Built with Tauri for native performance, just 10MB. Completely undetectable in video calls, screen shares, and recordings.

The Problem

Many professionals need real‑time AI assistance during meetings or interviews, but existing tools either compromise privacy or are easily detected by video‑call platforms. This creates a trade‑off between usefulness and stealth, especially for users who cannot afford or do not want a closed‑source solution.

What This Does

Pluely is an open‑source, Tauri‑based desktop application (≈10 MB) that functions as a privacy‑first AI assistant. The frontend lives in src/ (153 files) written in React and TypeScript, while the native layer is implemented in Rust under src-tauri/. Key components include: src-tauri/src/main.rs – entry point that launches the Tauri window. src-tauri/src/db/main.rs – SQLite‑backed storage for chat history and system prompts (src-tauri/src/db/migrations/). src/lib/functions/ – Rust‑exposed API helpers for AI‑provider requests and speech‑to‑text handling. src/pages/app/components/speech/ – UI for recording, VAD, and result display.

The app can capture system audio, overlay UI, and inject AI‑generated responses without exposing the assistant’s presence in screen shares or recordings.

How To Use It

Download: Pre‑built installers are available for macOS (.dmg), Windows (.msi / .exe), and Linux (.deb, .rpm, .AppImage) from the project releases page. No local build is required for immediate use. Source build (if desired): The repository uses package.json (npm) for frontend dependencies and src-tauri/Cargo.toml for Rust binaries. Development typically requires Node.js, Rust (via rustup), and the Tauri CLI, but no npm run tauri dev script is documented in the repo. Configuration: No environment‑variable setup is required for the binary releases; the app ships with default AI‑provider keys embedded. Custom providers can be added via the UI under src/pages/dev/components/ai-configs/ and stored in src/lib/storage/ai-providers.ts.

Real‑World Use

During a virtual interview, a user launches Pluely and selects “Always‑on‑top” mode. The RecordingPanel (src/pages/app/components/speech/RecordingPanel.tsx) captures system audio, feeds it to the STT module (src-tauri/src/speaker/), and displays real‑time transcriptions. Simultaneously, the AI‑response function (src/lib/functions/ai-response.function.ts) queries the selected provider and injects concise answers into the conversation window, all while remaining invisible to the meeting platform’s UI.

Code Health & Issues

No test files detected – repository‑wide untested code paths; adding unit tests for the Rust DB layer and React hooks would reduce risk. CI/CD present – GitHub Actions workflow .github/workflows/publish.yml handles releases, but no linting or test steps are defined. Database migrations exist (src-tauri/src/db/migrations/chat-history.sql, system-prompts.sql) but have no accompanying verification scripts. Dependencies – package-lock.json and Cargo.lock lock third‑party versions; a periodic security audit (e.g., npm audit, cargo audit) is recommended.

No other critical defects were found in the inspected structure.

The Bottom Line

Pluely delivers a lightweight, privacy‑oriented AI assistant that can run unobtrusively during calls, making it attractive for users who need real‑time help without sacrificing stealth. The lack of tests and CI‑integrated quality gates means it is best suited for early adopters or teams comfortable manually auditing the codebase; organizations requiring out‑of‑the‑box reliability may want to wait for a more mature release cycle.