Cognithor · Agent OS

Features

🧠

15 LLM Providers

Ollama (local), OpenAI, Anthropic, Gemini, Groq, DeepSeek, Mistral, Together, OpenRouter, xAI, Cerebras, GitHub Models, AWS Bedrock, Hugging Face, Moonshot. Auto-detects from API key.

📡

17 Channels

CLI, Web UI, REST API, Telegram, Discord, Slack, WhatsApp, Signal, iMessage, Teams, Matrix, Google Chat, Mattermost, Feishu, IRC, Twitch, Voice (STT/TTS). Auto-detect from tokens.

💾

5-Tier Cognitive Memory

Core identity, episodic logs, semantic knowledge graph, procedural skills, working memory. 3-channel hybrid search: BM25 + vector + graph.

🛡️

Enterprise Security

Deterministic Gatekeeper (4 risk levels), multi-level sandbox, SHA-256 audit chain, encrypted credential vault, EU AI Act compliance, red-teaming suite.

🎛️

React Control Center

Full web dashboard (React 19 + Vite 7) with backend launcher, live config editing, agent management, prompt editing, cron jobs, MCP servers, A2A settings.

🔧

MCP + A2A Protocols

13+ MCP tool servers (filesystem, shell, memory, web, browser, media). Agent-to-Agent protocol (Linux Foundation RC v1.0) for inter-agent communication.

⚙️

PGE Architecture

Planner (LLM) → Gatekeeper (deterministic policy) → Executor (sandboxed). No hallucinated tool calls — every action is policy-checked before execution.

🔒

100% Local-First

All data stays on your machine. No cloud required — Ollama runs fully local. Optional cloud LLM providers available. Full GDPR compliance.

📚

Procedural Learning

Reflector auto-synthesizes reusable skills from successful sessions. Learned procedures are suggested for future similar requests.

Architecture

┌──────────────────────────────────────────────────────────────┐
│           Control Center UI (React 19 + Vite 7)              │
├──────────────────────────────────────────────────────────────┤
│          REST API (FastAPI, 20+ endpoints, port 8741)        │
├──────────────────────────────────────────────────────────────┤
│                      Channels (17)                           │
│  CLI · Web · Telegram · Discord · Slack · WhatsApp · Signal  │
├──────────────────────────────────────────────────────────────┤
│                    Gateway Layer                              │
│            Session Management · Agent Loop                    │
├────────────┬──────────────┬──────────────────────────────────┤
│  Planner   │  Gatekeeper  │  Executor                        │
│  (LLM)     │  (Policy)    │  (Sandbox)                       │
├────────────┴──────────────┴──────────────────────────────────┤
│              MCP Tool Layer (13+) + A2A Protocol              │
├──────────────────────────────────────────────────────────────┤
│              Multi-LLM Backend Layer (15)                     │
├──────────────────────────────────────────────────────────────┤
│              5-Tier Cognitive Memory                          │
│  Core · Episodic · Semantic · Procedural · Working            │
└──────────────────────────────────────────────────────────────┘

Quick Start

# Clone and install
git clone https://github.com/Alex8791-cyber/cognithor.git
cd cognithor
pip install -e ".[all,dev]"

# Pull Ollama models
ollama pull qwen3:32b
ollama pull qwen3:8b
ollama pull nomic-embed-text

# Start CLI
cognithor

# Or start the Control Center UI
cd ui && npm install && npm run dev