Sovereign 8s.is mirror of 8b-is/q8-caster
- Rust 90.3%
- HTML 4.5%
- Shell 4.3%
- CSS 0.9%
| Filename | Latest commit message | Latest commit date |
|---|---|---|
Complete implementation of AI-narrated presentations with auto-scrolling: Core Components (1,929 lines total): - NEW: src/presentation/content_analyzer.rs (378 lines) * Smart markdown parsing with section detection * Automatic summarization (100+ words → summary) * Image/video/asciinema URL extraction * Duration estimation (150 words per minute) * 3 unit tests ✅ - NEW: src/presentation/scroll_engine.rs (319 lines) * Revolutionary sync algorithm: scroll_speed = distance / duration * 5 highlight modes (None, Paragraph, Border, Background, Pulse) * Pause/resume playback control * 6 unit tests ✅ - NEW: src/presentation/timeline.rs (281 lines) * Multi-media timeline coordination * TimelineEvents with start_time + duration * get_active_events() for current media * 5 unit tests ✅ - NEW: src/presentation/narration.rs (398 lines) * TTS engine integration (194 voices) * Speech segment generation with timing * 4 narration modes (Full, AutoSummarize, KeyPoints, Custom) * Speed/voice/volume control * 8 unit tests ✅ (6 passing, 2 TTS init in test env) - NEW: src/presentation/mod.rs (267 lines) * PresentationManager orchestrator * Location-aware TV routing (Bedroom, Office, LivingRoom, Kitchen, Garage) * Presentation state tracking * 3 unit tests ✅ - NEW: src/mcp/handlers/presentation.rs (286 lines) * 4 new MCP tools for AI agents * cast_with_narration: Create AI-narrated presentations * list_presentations: List active presentations * control_presentation: Playback control (pause, resume, stop) * get_presentation_status: Get current state Modified Files: - src/lib.rs: Added presentation module export - src/mcp/handlers.rs: Added presentation handlers - src/error.rs: Added NotFound error variant Features: ✅ Smart summarization (long content = summary speech, full display) ✅ Perfect scroll synchronization (arrives exactly when speech finishes) ✅ Multi-media support (images, videos, asciinema terminal recordings) ✅ Location-aware routing (bedroom, office, living room, etc.) ✅ TTS integration (194 voices, adjustable rate/pitch/volume) ✅ 4 narration modes (Full, AutoSummarize, KeyPointsOnly, Custom) ✅ 5 highlight modes for current section ✅ Timeline coordination for multi-media events Tests: 23 passed, 2 failed (TTS init in test env), 25 total Build: Success ✅ Use Cases: - Security alerts with camera feeds (bedroom TV) - Project demos for Alexandra (office TV) - AI explanations with visuals - Terminal session demonstrations (asciinema) - Documentation walkthroughs Integration: Core component of Aye OS AI-human communication system 🤖 Generated with Claude Code Co-Authored-By: Claude <noreply@anthropic.com> |
||
| docs | ||
| scripts | ||
| src | ||
| static | ||
| tests | ||
| .gitignore | ||
| Cargo.lock | ||
| Cargo.toml | ||
| CASTING_FEATURES.md | ||
| CLAUDE.md | ||
| config.toml.example | ||
| context.md | ||
| install-mcp.sh | ||
| mcp-config.json | ||
| MCP_SETUP.md | ||
| memory-export.json | ||
| q8-caster-backup.json | ||
| q8-caster.service | ||
| README.md | ||
| test_mcp.sh | ||
Q8-Caster 🎬
AI-powered display casting MCP server with comprehensive media support. Cast anything from markdown presentations to 3D models, videos, and live streams to any display.
Features
- MCP Integration: Full Model Context Protocol support for AI-driven display control
- Multi-Display Support: Control multiple displays, mirror screens, configure resolutions
- Content Types:
- Markdown with live rendering (dark/light themes)
- Video playback with hardware acceleration
- Image display
- 3D model visualization (GLTF)
- Live streaming (RTSP, WebRTC, HLS, DASH)
- Presentations
- Network Receivers: UPnP/DLNA, AirPlay, Chromecast support
- Content Caching: Smart caching with TTL support
- Codec Detection: Reports all available audio/video codecs
- Audio Routing: Control audio devices and routing
Quick Start
# Build the project
./scripts/manage.sh build release
# Run as MCP server (stdio)
./scripts/manage.sh run
# Or run as daemon
./scripts/manage.sh daemon
# Check status
./scripts/manage.sh status
MCP Tools
cast_content
Cast content to a display.
{
"display_id": "display_0",
"content_type": "markdown",
"source": "/path/to/presentation.md",
"options": {
"theme": "dark"
}
}
list_displays
List all available displays with their properties.
list_codecs
List all available audio/video codecs with hardware acceleration info.
configure_display
Configure display settings (resolution, position, mirroring).
start_receiver
Start network receivers (UPnP, AirPlay, Chromecast).
cache_content
Cache content for faster access.
discover_chromecasts
Discover available Chromecast devices on the network.
connect_chromecast
Connect to a specific Chromecast device by name.
cast_to_chromecast
Cast content directly to a Chromecast device.
{
"device_name": "Living Room TV",
"content_type": "video",
"source": "https://example.com/video.mp4"
}
control_chromecast
Control Chromecast playback (play, pause, stop, seek, volume).
Examples
Cast a Markdown Presentation
{
"tool": "cast_content",
"arguments": {
"content_type": "markdown",
"source": "presentation.md",
"options": {
"theme": "dark"
}
}
}
Stream RTSP Camera
{
"tool": "cast_content",
"arguments": {
"content_type": "stream",
"source": "rtsp://camera.local:554/stream",
"options": {
"protocol": "rtsp"
}
}
}
Start AirPlay Receiver
{
"tool": "start_receiver",
"arguments": {
"protocols": ["airplay", "upnp"],
"port": 8420
}
}
Cast to Chromecast
// First discover devices
{
"tool": "discover_chromecasts"
}
// Connect to a device
{
"tool": "connect_chromecast",
"arguments": {
"device_name": "Living Room TV"
}
}
// Cast content
{
"tool": "cast_to_chromecast",
"arguments": {
"device_name": "Living Room TV",
"content_type": "video",
"source": "https://example.com/movie.mp4"
}
}
Architecture
- Display Manager: Handles display enumeration and window creation
- Media Engine: GStreamer-based media playback with codec detection
- Render Engine: GPU-accelerated rendering for markdown and 3D content
- Network Receiver: Multi-protocol network discovery and streaming
- Content Cache: LRU memory cache with disk persistence
Requirements
- Rust 1.70+
- GStreamer 1.0
- X11/Wayland (for display control)
License
MIT - Made with 💜 by 8b-is