Documentation

Everything you need to use and operate CorvinOS

Operator Handbook

Complete visual guide to the CorvinOS console — every page explained with dark-mode screenshots and real data.

Open Handbook →
Handbook chapters
🗂️

Getting Started

Open the console, understand the Dashboard overview — engine status, messaging channels, audit log, and API key health at a glance.

Read the guide →
💬

Messaging

Chat directly from the browser, connect Discord / WhatsApp / Telegram / Slack channels, configure voice TTS, and manage the user directory.

Explore →
🤖

Intelligence

Select and configure AI engines (Claude Code, Hermes/Ollama, Copilot, Codex), customise personas with engine pinning and model overrides, and manage tenant files.

Learn more →
🔨

Build

Create multi-step workflows, monitor pipeline runs with DAG views, forge sandboxed tools, manage skills and MCP connectors, and set up Agent Hub endpoints.

Build guide →
🗄️

Data

Connect PostgreSQL, MySQL, S3, BigQuery, Snowflake, and HTTP data bridges. Set up RAG knowledge providers and browse the Knowledge Hub marketplace.

Data guide →
⚙️

System

Manage API keys (BYOK, RSA-OAEP encrypted), license, compliance audit chain, organisations, engine control matrix, LDD quality layers, and tenant settings.

System guide →
Core concepts
🔐

Security & Compliance

EU AI Act 2026 and GDPR as structural design constraints. Hash-chained audit log, consent gate, path-gate hook, bot-disclosure, and secret vault — all always-on.

Audit & Compliance →

Engine Architecture

Five engines — Claude Code, OpenCode, Hermes (local), Codex CLI, GitHub Copilot — with different roles, locality, and cost models. Swap engines per persona or per chat.

AI Engine guide →
📊

Loss-Driven Development

12-layer quality framework built into CorvinOS. Each layer is independently toggleable. Enforces reproducibility, E2E-first iteration, and docs-as-definition-of-done.

Quality (LDD) →
← Back to Home