Skip to content

nexsusagent-coder/SENTIENT_CORE

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

122 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

SENTIENT OS

🐺 SENTIENT OS

The Operating System That Thinks

Düşünen İşletim Sistemi

KurulumHızlı BaşlangıçLLM HubÖzelliklerMimariKullanım

Version Rust License Crates Models Providers Skills


📊 Proje Özeti

Metrik Değer Açıklama
Rust Crate 93 Tamamen Rust ile yazılmış
LLM Provider 57+ OpenAI, Anthropic, Google, DeepSeek, Groq, Unify, Portkey...
LLM Model 245+ GPT-2 (2019) → o4-mini/Grok 4 (2026)
AI Gateway 11 OpenRouter, Unify, Portkey, Helicone, NotDiamond...
Skill 5,587+ Dünyanın en büyük AI skill koleksiyonu
Entegrasyon 72+ proje AutoGPT, CrewAI, LangChain, Ollama...
Kanal 24 platform Telegram, Discord, WhatsApp, Slack...
Rust Kodu 152,877+ satır Production-grade

🚀 Tek Komutla Kurulum

OpenClaw Tarzı — Tek Komut, Tam Kurulum

# Yöntem 1: curl ile (önerilen)
curl -fsSL https://raw.githubusercontent.com/nexsusagent-coder/SENTIENT_CORE/main/install.sh | bash

# Yöntem 2: git clone + install
git clone https://github.com/nexsusagent-coder/SENTIENT_CORE.git
cd SENTIENT_CORE
./install.sh

Kurulum Adımları (OpenClaw Standardı)

1. ⚠️  Yasal Uyari (Yes/No)          → Kullanım koşulları
2. 🔍 Sistem Tespiti                → OS, RAM, GPU, CPU
3. 📦 Kurulum Modu                  → Quick / Full / Custom
4. 🧠 LLM Provider Seçimi          → Lokal (ücretsiz) veya API
5. 🧩 Ek Modüller                   → Voice, Browser, Desktop, Channels
6. 🐳 Docker Servisleri             → PostgreSQL, Redis, Qdrant...
7. ⚙️  Yapılandırma (.env)          → Otomatik oluşturma
8. 🔨 Derleme                       → cargo build --release
9. ✅ Doğrulama                     → health-check

Hızlı Kurulum (Sorusuz)

# Tüm soruları atla, varsayılanlarla kur
./install.sh --yes --quick

# Tam kurulum (tüm 93 crate + Docker)
./install.sh --yes --full

# Özel dizine kur
./install.sh --dir /opt/sentient

# Kaldır
./install.sh --uninstall

⚡ Hızlı Başlangıç

# İnteraktif sohbet
sentient chat

# Tek soru sor
sentient ask "Rust'ta async nasıl çalışır?"

# Sesli asistan (JARVIS modu)
sentient voice

# Web API sunucusu
sentient gateway
# → http://localhost:8080

# Web dashboard
sentient dashboard
# → http://localhost:8080/dashboard

# Sistem durumu
sentient status

# İlk kurulum sihirbazı
sentient init

# Sağlık kontrolü
sentient doctor

🧠 LLM Hub

SENTIENT OS dünyanın en kapsamlı LLM hub'ına sahiptir. 57+ provider, 245+ model desteği.

Doğrudan AI Şirketleri

Provider Modeller Fiyat Ücretsiz
OpenAI GPT-4o, o3, o4-mini (17 model) $$
Anthropic Claude 4, 3.5, 3, 2, 1 (12 model) $$
Google Gemini 2.5, 2.0, 1.5, Gemma 3 (14 model) $ ✅ Flash ücretsiz
Mistral 🇫🇷 Large 2, Small 3.1, Codestral, Pixtral (11 model) $ ✅ Bazı modeller
DeepSeek 🇨🇳 V3, R1, R2, V4, Coder (6 model) EN UCUZ ✅ Çoğu ücretsiz
xAI Grok 3, 3 Mini, Grok 4 (4 model) $$ ✅ Mini ücretsiz
Cohere Command A, R+, Aya Exa (6 model) $
Perplexity Sonar, Deep Research (4 model) $$

AI Gateway / Router (11 Provider)

Gateway İşlev Ücretsiz Tier
OpenRouter 300+ model marketplace ✅ Bazı modeller
Unify AI ML-based akıllı routing (quality/cost/speed) ✅ $5 kredi
Portkey Enterprise gateway, failover, caching ✅ 10K req/mo
Helicone AI observability, cost tracking ✅ 50K req/mo
NotDiamond ML ile prompt bazlı model seçimi
AI/ML API 100+ model, doğrudan erişimden %40 ucuz ✅ 100 req/gün
Glama Multi-model gateway + MCP desteği
Requesty LLM router, A/B testing
LiteLLM 100+ provider, self-hosted proxy Açık kaynak
Cloudflare Workers AI Edge inference
Chutes Tamamen ücretsiz inference! ✅ Tamamen

Lokal / Açık Kaynak (Ücretsiz!)

Model VRAM Ollama ID Özellik
Llama 4 Scout 48GB llama4:scout 10M context, MoE 109B, Vision
Llama 4 Maverick 96GB llama4:maverick 1M context, MoE 400B, Vision
Qwen3 30B-A3B MoE 4GB! qwen3:30b-a3b 3B aktif parametre, Reasoning
DeepSeek R2 96GB deepseek-r2 Reasoning + Vision
Gemma 3 27B 16GB gemma3:27b Vision, 128K context
Phi-4 14B 8GB phi4:14b MIT lisans, Code
Mistral Small 3.1 16GB mistral-small3.1:24b Vision + Tools
Gemma 3 4B 4GB gemma3:4b Vision, en hafif
DeepSeek R1 Distill 8B 4GB deepseek-r1:8b En küçük reasoning
Qwen 2.5 Coder 14B 8GB qwen2.5-coder:14b Code specialist

Çin AI Ekosistemi

Provider Modeller
Qwen (Alibaba) Qwen3, QwQ, QVQ, Qwen4 Max
Baidu ERNIE ERNIE 4.0, 3.5
Zhipu GLM GLM-4 Plus, Flash
Moonshot (Kimi) V1 128K
StepFun Step-2, Step-1V
ByteDance Doubao 1.5 Pro

Diğer Bölgesel AI

Bölge Provider Modeller
🇷🇺 Rusya GigaChat Pro, Max
🇰🇷 Kore Upstage Solar Pro 2
🇪🇺 Avrupa Aleph Alpha Luminous, Pharia
🇮🇳 Hindistan Sarvam AI Sarvam-M
🇯🇵 Japonya Rinna, CyberAgent CALM 3
🇦🇪 Arap InceptionAI JAIS 30B

Rust ile Kullanım

use sentient_llm::{LlmHub, ChatRequest, Message};

#[tokio::main]
async fn main() {
    // Tüm provider'ları otomatik algıla (.env dosyasından)
    let hub = LlmHub::from_env().unwrap();

    // Sohbet
    let response = hub.chat(ChatRequest {
        model: "gpt-4o".into(),
        messages: vec![Message::user("Merhaba!")],
        ..Default::default()
    }).await.unwrap();

    println!("{}", response.choices[0].message.content.as_text().unwrap());
}

🌟 Özellikler

🧠 Core Intelligence

Özellik Açıklama
57+ LLM Provider OpenAI, Anthropic, Google, Mistral, DeepSeek, Groq, Unify, Portkey...
245+ Model GPT-2 (2019) → o4-mini/Grok 4 (2026)
11 AI Gateway OpenRouter, Unify, Portkey, Helicone, NotDiamond...
Smart Routing Maliyet, hız veya kalite bazlı otomatik provider seçimi
Circuit Breaker Hata toleransı, otomatik retry ve failover
Cost Tracker Gerçek zamanlı maliyet takibi

🎙️ Voice (JARVIS Seviyesi)

🎤 Mikrofon → Whisper STT → LLM → TTS → 🔊 Hoparlör
     ↑                                    |
     └──── Wake Word ("Hey Sentient") ────┘
Özellik Lokal (Ücretsiz) API (Ücretli)
STT Whisper.cpp OpenAI Whisper
TTS Piper (Türkçe) ElevenLabs
Wake Word Porcupine, Vosk -

🖥️ Desktop Automation (Computer Use)

Özellik Açıklama
Fare/Klavye İnsan gibi hareket (Bezier eğrileri)
Ekran Okuma OCR, görüntü analizi
GUI Kontrol Tüm uygulamalar
Safety 50+ tehlikeli komut engeli

📱 24 Platform Entegrasyonu

Platform Durum Tür
Telegram ✅ Aktif Mesajlaşma
Discord ✅ Aktif Mesajlaşma
WhatsApp ✅ Aktif Mesajlaşma
Slack ✅ Aktif İş
Email ✅ Aktif İletişim
+ 19 platform daha

🔐 Güvenlik (Enterprise Grade)

Özellik Açıklama
V-GATE API key'ler asla istemcide değil, proxy üzerinden
Guardrails Prompt injection, PII, secret filtreleme
Vault AES-256-GCM şifreleme, key rotation
TEE AMD SEV-SNP, Intel TDX desteği
ZK-MCP Zero-knowledge proof
Audit Tüm işlemler loglanır

📦 Crate Yapısı (93 Crate)

🧠 Core (7 Crate)

Crate İşlev
sentient_core Ana motor, sistem yönetimi
sentient_memory Agent bellek sistemi (Hippocampus)
sentient_graph Event graph, workflow
sentient_orchestrator Agent loop & routing
sentient_gateway API gateway
sentient_cevahir Türkçe LLM cognitive engine
oasis_brain Gemma 4 cognitive kernel

🤖 LLM (7 Crate)

Crate İşlev
sentient_llm 57+ provider, 245+ model hub
sentient_embed Multi-provider embedding hub
sentient_rerank Search result reranking
sentient_local Lokal LLM yönetimi
sentient_groq Groq LPU ultra-fast inference
sentient_vector Vector search
sentient_rag Native RAG engine

🔒 Security (6 Crate)

Crate İşlev
sentient_vgate V-GATE proxy (API key güvenliği)
sentient_guardrails Input/output filtreleme
oasis_vault Secrets manager (AES-256-GCM)
sentient_tee TEE support (AMD SEV-SNP)
sentient_zk_mcp Zero-knowledge proof
sentient_anomaly Anomaly detection

🖥️ Desktop & Browser (5 Crate)

Crate İşlev
oasis_hands Desktop automation (EN BÜYÜK - 36K+ satır)
oasis_autonomous Tam otonom agent
oasis_browser Browser automation
oasis_manus Docker execution sandbox
sentient_desktop Computer Use / GUI control

🎤 Voice (2 Crate)

Crate İşlev
sentient_voice STT + TTS pipeline
sentient_wake Wake word detection

📱 Channels & Communication (1 Crate)

Crate İşlev
sentient_channels 24 platform entegrasyonu

🔧 Tools & Utilities (60+ Crate)

Crate İşlev
sentient_rag Native RAG engine
sentient_vision Vision/multimodal
sentient_mcp Model Context Protocol
sentient_sandbox E2B sandbox
sentient_skills Skills system (5,587+)
sentient_search Web search (Tavily, Brave, SearXNG)
sentient_image Image generation (DALL-E, SD, Flux)
sentient_video Video generation
sentient_finetuning Fine-tuning (LoRA)
sentient_quantize Model quantization
sentient_knowledge Knowledge graph
sentient_enterprise RBAC, SSO, Audit
sentient_compliance SOC 2 compliance
sentient_observability OpenTelemetry, Prometheus
sentient_i18n 8 dil desteği
sentient_backup Backup & DR
sentient_cluster Kubernetes operator
sentient_web Web server
sentient_workflow Visual flow builder (n8n-style)
sentient_daemon Background always-on assistant
sentient_proactive Time/Event/Pattern triggers
sentient_connectors Gmail, Calendar, GitHub, Weather
sentient_digest Morning briefing system
sentient_a2a Agent-to-Agent protocol
sentient_home Smart Home (Home Assistant)
sentient_social Social media automation
sentient_remote Mobile PWA, Telegram Mini App
sentient_learning User behavior learning
sentient_marketplace Skills marketplace
... ve 30+ crate daha

🎯 Kullanım Rehberi

1. İnteraktif Sohbet (REPL)

sentient chat
# veya model belirterek
sentient chat --model gpt-4o
sentient chat --model ollama:qwen3:30b-a3b

2. API Sunucusu (Gateway)

sentient gateway
# POST http://localhost:8080/v1/chat/completions
# OpenAI-compatible API
curl http://localhost:8080/v1/chat/completions \
  -H "Authorization: Bearer $JWT_SECRET" \
  -d '{"model":"gpt-4o","messages":[{"role":"user","content":"Merhaba"}]}'

3. Sesli Asistan

sentient voice
# "Hey Sentient" → mikrofon dinler → yanıt verir

4. Desktop Automation

sentient desktop
# Bilgisayarı otonom kontrol eder

5. Skill Kullanımı

sentient skill search "browser automation"
sentient skill run code-review --path ./src
sentient skill list

6. Docker ile Tüm Servisler

docker-compose up -d
# Servisler:
# - PostgreSQL (5432)    → Veritabanı
# - Redis (6379)         → Cache
# - Qdrant (6333)        → Vector DB
# - MinIO (9000/9001)    → Object Storage
# - Prometheus (9090)    → Monitoring
# - Grafana (3001)       → Dashboard
# - Ollama (11434)       → Lokal LLM
# - SearXNG (8888)       → Arama Motoru
# - RabbitMQ (5672)      → Message Queue

7. Sağlık Kontrolü

sentient doctor
# veya
./scripts/health-check.sh

8. Sistem Durumu

sentient status

📊 Sistem Gereksinimleri

Mod RAM Disk GPU Açıklama
API Only 2 GB 5 GB Yok API provider kullanımı
Local Small 8 GB 10 GB Opsiyonel Ollama 3B-7B
Local Medium 16 GB 20 GB 8GB VRAM Ollama 14B-27B (önerilen)
Local Large 64 GB 50 GB 24GB VRAM Ollama 70B+
Full Stack 32 GB 30 GB 8GB VRAM Tüm servisler + Docker
Enterprise 64 GB+ 100 GB+ 24GB+ VRAM Multi-tenant, RBAC

🐳 Docker

# Tüm servisleri başlat
docker-compose up -d

# Sadece temel servisler
./scripts/start.sh --minimal

# Servisleri durdur
./scripts/stop.sh

🛠️ Geliştirme

# Tüm crate'leri derle
cargo build --release

# Sadece CLI binary
cargo build --release --bin sentient

# Test
cargo test --workspace

# Clippy
cargo clippy --workspace

# Format
cargo fmt

# Belirli crate derle
cargo build --release -p sentient_llm
cargo build --release -p sentient_voice

Makefile Komutları

make build          # Derle
make run            # REPL başlat
make test           # Test çalıştır
make docker-up      # Docker servisleri başlat
make clean          # Temizle
make skills         # Skill library güncelle
make help           # Tüm komutlar

🌍 Dil Desteği

Dil Durum
🇹🇷 Türkçe ✅ Tam destek (Cevahir AI engine)
🇺🇸 English ✅ Full support
🇩🇪 Deutsch 🔄 Partial
🇫🇷 Français 🔄 Partial
🇪🇸 Español 🔄 Partial
🇯🇵 日本語 🔄 Partial
🇨🇳 中文 🔄 Partial
🇰🇷 한국어 🔄 Partial

📁 Proje Yapısı

SENTIENT_CORE/
├── crates/                  # 93 Rust crate
│   ├── sentient_core/       # Ana motor
│   ├── sentient_llm/        # LLM hub (57+ provider, 245+ model)
│   ├── sentient_embed/      # Embedding hub
│   ├── sentient_rerank/     # Reranking engine
│   ├── sentient_voice/      # Ses modülü (STT + TTS)
│   ├── sentient_channels/   # 24 platform
│   ├── sentient_memory/     # Bellek sistemi
│   ├── sentient_rag/        # RAG engine
│   ├── sentient_vision/     # Vision/multimodal
│   ├── sentient_mcp/        # MCP protocol
│   ├── sentient_vgate/      # V-GATE proxy
│   ├── sentient_guardrails/ # Güvenlik duvarı
│   ├── sentient_cli/        # CLI arayüzü
│   ├── sentient_setup/      # Setup wizard (TUI)
│   ├── oasis_hands/         # Desktop automation (36K+ satır)
│   ├── oasis_autonomous/    # Tam otonom agent
│   ├── oasis_browser/       # Browser automation
│   ├── sentient_cevahir/    # Türkçe LLM cognitive engine
│   └── ...                  # 75+ crate daha
├── integrations/            # 72+ entegre proje
│   ├── agents/              # AutoGPT, CrewAI, LangChain, Goose...
│   ├── framework/           # LlamaIndex, Phidata, Semantic Kernel...
│   ├── tools/               # Firecrawl, Mem0, Crawl4AI...
│   ├── browser/             # Browser-use, LightPanda...
│   ├── memory/              # Qdrant, MemGPT...
│   ├── sandbox/             # E2B, LocalStack...
│   ├── cevahir_ai/          # Türkçe LLM engine
│   └── security/            # NeMo-Guardrails...
├── skills/                  # 5,587+ skill
│   ├── Dev/                 # 2,965 skill
│   ├── OSINT/               # 1,050 skill
│   ├── Social/              # 238 skill
│   ├── Automation/          # 306 skill
│   ├── Media/               # 246 skill
│   └── ...                  # 5 kategori daha
├── scripts/                 # Yardımcı scriptler
│   ├── start.sh             # Tüm servisleri başlat
│   ├── stop.sh              # Servisleri durdur
│   ├── health-check.sh      # Sağlık kontrolü
│   ├── setup-ollama.sh      # Ollama kurulumu
│   └── run_tests.sh         # Test çalıştırıcı
├── data/                    # Veritabanları
├── dashboard/               # Web dashboard
├── config/                  # Konfigürasyon
├── docker-compose.yml       # 9 Docker servisi
├── install.sh               # ✨ Tek komutla kurulum (OpenClaw-style)
├── Makefile                 # Build komutları
├── .env.example             # Config şablonu
└── Cargo.toml               # Workspace config (93 member)

🔌 Entegrasyonlar (72+ Proje)

Agent Framework'leri (17)

Proje Açıklama
AutoGPT Otonom agent
CrewAI Multi-agent orchestration
LangChain LLM framework
LlamaIndex RAG framework
MetaGPT Multi-agent simulator
AutoGen Microsoft multi-agent
OpenHands AI developer
Goose AI coding agent
Phidata AI apps
Semantic Kernel Microsoft AI

Tools (5)

Proje Açıklama
Firecrawl Web scraping
Mem0 Memory layer
CAG4AI Web extraction
Judge0 Code execution
RAGFlow RAG engine

Türkçe LLM — Cevahir AI

Modül İşlev
Neural Network (V-7) RoPE, RMSNorm, SwiGLU, KV Cache, MoE, GQA
Cognitive Strategies Direct, Think, Debate, Tree of Thoughts
Turkish BPE Tokenizer 60K vocabulary, GPU batch processing
Memory & RAG Vector store, semantic cache

📜 Lisans

GNU AGPL v3.0 — Kullan, değiştir, paylaş.

Ticari kullanım için: enterprise@sentient-os.ai


🤝 Katkıda Bulunma

  1. Fork'la
  2. Branch oluştur: git checkout -b feature/yeni-ozellik
  3. Commit'le: git commit -m 'Yeni özellik'
  4. Push'la: git push origin feature/yeni-ozellik
  5. Pull Request aç

📞 İletişim

Kanal Link
GitHub github.com/nexsusagent-coder/SENTIENT_CORE
Ko-fi ko-fi.com/sentientos
Email sentient@sentient-os.ai

SENTIENT OS
The Operating System That Thinks

Made with 🦀 by the SENTIENT Team

About

🧠 SENTIENT OS - The Operating System That Thinks | Autonomous, Secure, High-Performance AI OS with 100% Native Rust Core

Topics

Resources

License

Code of conduct

Contributing

Security policy

Stars

Watchers

Forks

Packages

 
 
 

Contributors