Supreme Architect of The Code Eternal. An advanced autonomous AI agent system with cognitive architecture (Orythix), local multi-model routing, self-healing incident response, and high-performance terminal interface.

VIKI (Virtual Intelligence Knowledge Interface) is a production-grade, self-hosted autonomous AI agent and developer-first AI engineer written in Python. Built for local-first privacy and zero data leakage, VIKI runs natively on your hardware via LM Studio and Ollama, while seamlessly supporting cloud frontier models (OpenAI GPT-4o, Anthropic Claude 3.7 Sonnet, Google Gemini 2.5 Pro).

Powered by the Orythix Cognitive Architecture, VIKI provides deep ReAct reasoning, multi-agent swarm orchestration, automated incident healing (Sentry/Datadog stack trace ingestion), persistent SQLite RAG memory, and self-evolution through the Neural Forge—all accessible via an ultra-fast CLI-first REPL and web dashboard.

VIKI（虚拟智能知识接口）是一个高性自托管自主 AI 智能体与开发者优先的 AI 工程师系统。VIKI 专为本地优先与隐私安全设计，能够在无数据泄漏的前提下在本地硬件上运行（基于 LM Studio 和 Ollama），同时也无缝支持云端顶级大模型（OpenAI GPT-4o、Anthropic Claude 3.7 Sonnet、Google Gemini 2.5 Pro）。

• 100% 本地与物理隔离: 默认操作无需依赖外网 API（开启 VIKI_AIR_GAP=1 模式），保障核心代码与数据安全。
• Orythix 认知架构: 具备五层认知大脑（感知、判断、思考、反思、执行与元认知），在执行前自动审计计划，有效降低幻觉。
• 多模型灵活路由: 自动分发任务至本地模型（LM Studio/Ollama）或云端模型（OpenAI/Claude/Gemini）。
• 全生命周期软件工程与 Jira 自动化: 自动解析 Jira 需求、生成技术规范、编写 TDD 测试套件、执行 AST 代码重构以及生成 OpenAPI/gRPC 接口。
• 自主故障修复: 支持对接 Sentry/Datadog 告警，在隔离的 Git Worktree 中重现 Bug 并自动修复。
• 神经铸造厂 (Neural Forge): 无需 GPU 训练，将交互经验与 SQLite 记忆直接“烘焙”到本地模型的 System Prompt 中。

• 100% Sovereign & Air-Gap Capable: Default operation never leaves your local network (VIKI_AIR_GAP=1). Zero telemetry, zero external tracking.
• Orythix 5-Layer Cognitive Cortex: Judgment, Deliberation, Reflection, Execution, and Meta-Cognition layers reduce hallucinations and audit plans before execution.
• Multi-Model Provider Routing: Instantly switch or auto-route between LM Studio, Ollama, OpenAI, Anthropic Claude, and Google Gemini.
• Full-SDLC & Jira Automation: Automated Jira ticket parsing, technical spec generation, TDD test suites, AST codemods, and OpenAPI 3.1 / gRPC proto generation.
• Autonomous Incident Healing: Connects to Sentry/Datadog webhooks, reproduces bugs in isolated Git worktrees, applies targeted fixes, verifies via pytest, and creates PRs.
• Multi-Agent Swarm Engineering: Execute complex DAG tasks using leader-worker agent swarms and Monte Carlo Tree Search (MCTS) reasoning.
• Local Neural Forge: Bakes user feedback and lessons accumulated in SQLite directly into custom model system prompts without GPU training.
• Low-Hardware Optimization: Runs efficiently on 8 GB RAM laptops (with dedicated 4k token caps, prompt compression, and LRU entity extraction caching).
• Capability Gating & Security: Sandboxed file and shell execution (viki/core/sandbox.py) with strict permission boundaries and prompt injection detection.

VIKI dynamically routes queries to the optimal provider profile configured in config/models.yaml:

VIKI executes tasks through a deterministic 5-Layer Consciousness Cortex:

graph TD
    A["User Prompt / Event Trigger"] --> B["Layer 1: Perception & Request Pipeline"]
    B --> C["Layer 2: Judgment Engine (Reflex vs Deliberation)"]
    C -->|Reflex Path| D["Instant Short-Circuit Execution"]
    C -->|Deliberate Path| E["Layer 3: Cortex & Multi-Model Router"]
    E --> F["Layer 4: Reflection & Hallucination Audit"]
    F --> G["Layer 5: Execution & Capability Registry"]
    G --> H["Layer 6: Meta-Cognition & Memory Persistence"]
    H --> I["SQLite RAG Wisdom DB & Neural Forge"]

• Python 3.10+ (3.10, 3.11, and 3.12 supported)
• Local Inference Server: LM Studio or Ollama running locally (http://localhost:1234 or http://localhost:11434)

# Clone repository
git clone https://github.com/Orythix/viki.git
cd viki

# Create & activate virtual environment
python -m venv .venv
# On Windows (PowerShell):
./.venv/Scripts/Activate.ps1
# On Linux/macOS:
source .venv/bin/activate

# Install core package
pip install -e .

cp .env.example .env

(Optionally set OPENAI_API_KEY, ANTHROPIC_API_KEY, GEMINI_API_KEY, or custom VIKI_PERSONA in .env)

# Start interactive CLI REPL in current workspace
viki

# Single-shot command (runs task and exits)
viki "refactor auth logic in src/viki/core/security.py"

# Run with Dev Persona (coding-focused skills)
VIKI_PERSONA=dev viki

Run VIKI in an isolated container using Docker Compose:

# Build and run interactive container
docker compose build
docker compose run --rm -it viki

See docs/DOCKER.md for advanced container configurations.

VIKI includes 40+ built-in engineering playbooks spanning the complete software development lifecycle:

• Define & Plan: idea_refine, spec_driven_development, planning_and_task_breakdown, jira_sdlc_workflow
• Build & Refactor: incremental_implementation, test_driven_development, ast_codemod, openapi_schema, frontend_ui_engineering
• Verify & Review: browser_testing_with_devtools, debugging_and_error_recovery, code_review_and_quality, security_and_hardening
• Ship & Recover: git_workflow_and_versioning, ci_cd_and_automation, autonomous_incident_healing

Bake your accumulated SQLite lessons and project wisdom directly into custom LM Studio model profiles without GPU training:

# Bake top lessons into data/Modelfile.viki_evolved
python scripts/build_viki_model.py

# Set models.yaml default profile to lmstudio-gemma4e4b
python scripts/build_viki_model.py --set-default

• Full Documentation Sitemap
• Installation & Setup Guide
• VIKI Runbook & Troubleshooting
• Docker Deployment Guide
• System Architecture & Data Flow
• Security Policy & Capability Gating

Local AI Agent • Autonomous AI Engineer • Self-Hosted AI Assistant • LM Studio Agent • Ollama Agent • Air-Gapped AI • Sovereign AI • ReAct Reasoning • Model Context Protocol (MCP) • Agentic SDLC • Sentry Bug Healing • Neural Forge • Private LLM • Python AI Agent • 中文 AI 智能体 • 本地大模型

VIKI is licensed under the Apache License 2.0.