Files
paseo/packages/server
Mohamed BoudraandGitHub 48e65b407f Replace Parcel Watcher with Custom Watcher (#3107)
* fix(server): replace wedging file observation

Native watcher teardown could block the daemon main thread during workspace archive. Own recursive observation behind an awaited lifecycle boundary, retain an isolated comparison harness, and gate correctness, resources, and sustained CPU across platforms.

* test(server): accept root watcher evidence

Native recursive watchers may coalesce ignored descendant changes into a root event. Assert that ignored paths stay excluded while a nonignored sentinel proves delivery remains live.

* test(server): preserve watcher comparison evidence

Record Parcel delivery misses without weakening the production observer gate, and allow the Windows stress test to use its declared polling window.

* fix(server): reconcile native watcher omissions

Back macOS and Windows recursive watching with a bounded namespace inventory so coalesced create and delete events converge without unbounded rescans.

* test(server): compare worktree filesystem identity

Windows can return the same worktree through 8.3 and long path spellings. Preserve cardinality and branch checks while comparing the real path.

* perf(server): bound native inventory audits

Reduce full native inventory cadence while preserving a six-second convergence bound, and gate setup readiness and event-loop p99 against platform-appropriate budgets.

* build: refresh Nix npm dependency hash

* perf(server): scope native file reconciliation

* perf(server): defer native safety audits until quiet

* fix(server): move native safety deadline during audits

* perf(server): audit native inventories after settled churn

* fix(server): reconcile native root evidence promptly

* perf(server): deduplicate native audit scheduling

* perf(server): coalesce native event clusters

* perf(server): suppress fresh native change scans

* test(server): separate burst recovery from sustained churn

* test(server): stabilize observer performance gates

* perf(server): skip fresh native change scans

* fix(server): preserve nested native change audits

* perf(server): coalesce native repair scans

* fix(server): reconcile native directory moves promptly

* test(server): bound observer burst repair cost

* refactor(server): own file observer lifecycle

* build: refresh Nix npm dependency hash

* ci: fold observer coverage into server suite

* fix(server): await file observer shutdown
2026-08-10 13:43:48 +02:00
..
2026-05-28 01:58:18 +08:00
2026-04-23 21:44:48 +07:00

Voice Assistant

A voice-controlled terminal assistant that runs as a single local service.

Quick Start

# Install dependencies
npm install

# Copy environment variables
cp .env.example .env
# Edit .env and add your API keys (OpenAI, Deepgram)

# Run development servers
npm run dev

# Open browser to http://localhost:5173

Architecture

  • Express Server (port 3000) - Serves API and built UI in production
  • Vite Dev Server (port 5173) - Hot-reload React UI in development
  • WebSocket (/ws) - Real-time bidirectional communication
  • Agent - STT → LLM → TTS pipeline with terminal control
  • Daemon - tmux-based terminal management (in-process)

Development

# Run both servers (recommended)
npm run dev

# Or run separately:
npm run dev:server  # Express on port 3000
npm run dev:ui      # Vite on port 5173

# Type checking
npm run typecheck

# Build for production
npm run build

# Start production server
npm start

Project Status

✅ Completed (Phases 1-2):

  • Package setup and configuration
  • Express server with WebSocket
  • React UI with Vite
  • WebSocket client with ping/pong testing

⏳ In Progress (Phase 3):

  • Terminal control (tmux integration)

📋 Planned (Phases 4-9):

  • LLM integration (OpenAI GPT-4)
  • Agent orchestrator
  • Speech-to-Text (Deepgram)
  • Text-to-Speech (OpenAI)
  • Audio streaming
  • UI polish

See IMPLEMENTATION_PLAN.md for complete details.

Environment Variables

OPENAI_API_KEY=your-openai-key-here      # GPT-4 and TTS
DEEPGRAM_API_KEY=your-deepgram-key-here  # Streaming STT
STT_MODEL=whisper-1        # Optional: override to gpt-4o-transcribe, etc.
STT_CONFIDENCE_THRESHOLD=-3.0  # Optional: reject low-confidence clips
STT_DEBUG_AUDIO_DIR=.stt-debug # Optional: persist raw dictation audio for debugging
PASEO_HOME=~/.paseo        # Runtime state directory (agents/, etc.)
PASEO_LISTEN=127.0.0.1:6767  # Listen address (host:port or /path/to/socket)

PASEO_HOME defaults to ~/.paseo and isolates runtime artifacts like agents/. PASEO_LISTEN controls the daemon listen address. For blue/green testing you can run a parallel server without touching production state:

PASEO_HOME=~/.paseo-blue PASEO_LISTEN=127.0.0.1:7777 npm run dev

Tech Stack

  • Server: Express, TypeScript, ws (WebSocket)
  • Client: React 18, Vite, TypeScript
  • Terminal: tmux (via child_process)
  • AI: OpenAI (LLM + TTS), Deepgram (STT)

Testing

Currently manual testing via:

  1. Start servers: npm run dev
  2. Open http://localhost:5173
  3. Test WebSocket connection (green status indicator)
  4. Click "Send Ping" button to test communication

More testing guidance as features are implemented.

License

MIT