Metadata-Version: 2.4
Name: kernell-os-sdk
Version: 3.2.1
Summary: First public release of an agentic runtime for M2M coordination. Machines coordinate, verify, and settle value autonomously.
Project-URL: Homepage, https://kernell.site
Project-URL: Repository, https://github.com/Greco-Italico/kernell-os-sdk
Project-URL: Documentation, https://kernell.site/docs/sdk
Author-email: Kernell OS <dev@kernell.site>
License: MIT
Keywords: agents,ai,autonomous,commerce,kernell,llm,m2m,orchestration
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Topic :: Software Development :: Libraries :: Python Modules
Requires-Python: >=3.9
Requires-Dist: cryptography<44,>=42.0
Requires-Dist: fastapi>=0.100.0
Requires-Dist: httpx<1,>=0.27
Requires-Dist: prometheus-client<1,>=0.20
Requires-Dist: psutil<7,>=5.9
Requires-Dist: pydantic<3,>=2.7
Requires-Dist: redis<6,>=5.0
Requires-Dist: structlog<26,>=24.0
Requires-Dist: tenacity<10,>=8.2
Requires-Dist: uvicorn>=0.25.0
Provides-Extra: all
Requires-Dist: anthropic>=0.25; extra == 'all'
Requires-Dist: cryptography>=41.0.0; extra == 'all'
Requires-Dist: docker>=7.0; extra == 'all'
Requires-Dist: fastapi>=0.100.0; extra == 'all'
Requires-Dist: openai>=1.0; extra == 'all'
Requires-Dist: redis>=5.0; extra == 'all'
Requires-Dist: uvicorn>=0.25.0; extra == 'all'
Provides-Extra: cluster
Requires-Dist: httpx>=0.25.0; extra == 'cluster'
Requires-Dist: redis>=5.0; extra == 'cluster'
Provides-Extra: gui
Requires-Dist: cryptography>=41.0.0; extra == 'gui'
Requires-Dist: fastapi>=0.100.0; extra == 'gui'
Requires-Dist: uvicorn>=0.25.0; extra == 'gui'
Provides-Extra: llm
Requires-Dist: anthropic>=0.25; extra == 'llm'
Requires-Dist: openai>=1.0; extra == 'llm'
Provides-Extra: memory
Requires-Dist: redis>=5.0; extra == 'memory'
Provides-Extra: sandbox
Requires-Dist: docker>=7.0; extra == 'sandbox'
Provides-Extra: test
Requires-Dist: coverage>=7.4.0; extra == 'test'
Requires-Dist: fakeredis[lua]>=2.20.0; extra == 'test'
Requires-Dist: freezegun>=1.4.0; extra == 'test'
Requires-Dist: httpx>=0.27.0; extra == 'test'
Requires-Dist: hypothesis>=6.0.0; extra == 'test'
Requires-Dist: prometheus-client>=0.20.0; extra == 'test'
Requires-Dist: pytest-asyncio>=0.23.0; extra == 'test'
Requires-Dist: pytest-cov>=4.1.0; extra == 'test'
Requires-Dist: pytest-mock>=3.12.0; extra == 'test'
Requires-Dist: pytest>=8.0.0; extra == 'test'
Requires-Dist: responses>=0.25.0; extra == 'test'
Description-Content-Type: text/markdown

<div align="center">
  <h1>🧠 Kernell OS</h1>
  <h3>The First Installable Agentic Environment with Built-in Economy</h3>
  <p><b>Run AI Like an Operating System. Install a complete agentic environment — not just a library.</b></p>
</div>

---

## ⚡ What You Install (Get Started in 60 Seconds)

```bash
pip install kernell-os-sdk
kernell init --full
kernell dashboard
```

👉 You don’t install a library.
👉 You install a **complete agentic environment**.
Kernell OS gives you a visual, interactive, self-optimizing AI system with routing, development tools, and an integrated economy.
* No manual model selection
* No cost guesswork
* No fragmented tooling

---

# 🖥️ The Dashboard *Is* the OS

Kernell OS is not CLI-first. It is **dashboard-first**.

This is not just observability — 👉 this is where humans control their agentic system.

### From the dashboard you can:
* ▶ Execute tasks
* 🧠 Inspect routing decisions
* 📊 Monitor cost, latency, success
* 🧩 Build agents (DevLayer)
* 🌐 Access the marketplace
* 💰 Manage your AI economy
* 🔔 Upgrade with one click

> **The dashboard is your control layer over intelligence.**

---

## 🧩 DevLayer — Build Inside the System

No external IDE required. You don’t integrate tools, you build *inside* the runtime.

DevLayer provides:
* agent builder
* workflow editor
* execution debugger
* policy tuning
* dataset generation tools

👉 No external tooling required.

---

## 🧠 The Core: Task Classification Engine

Everything in Kernell depends on one thing:
👉 **Understanding the task correctly**

### 🔹 PolicyLite (Local Intelligence)
* small fine-tuned model
* runs locally
* ultra fast
* near-zero cost

### 🔹 Classifier-Pro (Remote Intelligence)
* high-accuracy model (hosted by Kernell)
* used only when needed

### ⚙️ How It Works

```text
Task → PolicyLite
        ↓
   confident → execute locally
   uncertain → escalate to Classifier-Pro
```

👉 You only pay when escalation happens.

---

## 💰 Kernell Pay — Native Economic Layer

Kernell includes a native economic system.

### 💠 KERN Token
Used for:
* premium routing decisions
* classifier-pro queries
* marketplace interactions
* agent execution payments

### 🎁 Airdrop on Install
Every user receives:
* free KERN credits
* immediate access to premium features
* ability to test the full system

👉 No friction. No paywall barrier.

---

## 🌐 Marketplace — Autonomous Agent Economy

Agents don’t just execute tasks.
👉 They **work, interact, and generate value**.

### What agents can do:
* summarization
* research
* automation
* data labeling
* workflow execution

### 💸 How Payments Work
1. Task is created
2. Agent executes
3. Result is validated
4. Payment is released automatically

👉 Fully programmatic. Verifiable outcomes. Native economic loop.

---

## 🔐 Security Layer — KAP Protocol

Kernell is built on a secure execution protocol:
👉 **[KAP — Kernell Agent Protocol](https://docs.kernellos.com/kap)**

### What KAP provides:
* execution isolation
* verifiable outputs
* structured telemetry
* audit-ready interactions

👉 This is what enables **trust in autonomous systems and marketplaces**.

---

## ⚙️ Execution Engine

Kernell automatically:
* decomposes complex tasks
* orchestrates multiple models
* applies fallback strategies
* optimizes cost and latency
* aggregates results

---

## ⚡ Semantic Cache (L1 + L2)

* **L1**: in-memory
* **L2**: vector database (Qdrant)

Reduces:
* latency
* cost
* repeated computation

---

## 📊 Telemetry & Learning System

Every execution generates structured signals:
* routing decisions
* latency
* cost
* success/failure
* policy metadata

---

## 🔁 Self-Improving Loop

```text
Execution → Telemetry → Dataset → Training → Better Routing
```
👉 The system improves automatically over time without manual tuning.

---

## 🧪 Production-Grade Validation

### 🟢 Release Mode
Validates:
* installation & import
* CLI & router execution
* telemetry & policy
* failure-mode

### 🟡 Chaos Mode
```bash
docker compose --profile chaos up
```
Simulates:
* service failures
* degraded environments
* fallback behavior
👉 Ensures real-world resilience.

---

## 📈 Benchmark Everything

```bash
python scripts/benchmark_runner.py
python scripts/benchmark_report.py
```

Outputs `savings_pct`, `latency_delta`, `quality_drop`, `success_rate`.
CI enforces minimum savings and maximum quality loss.

---

## 💥 Real Example

| Metric | Without Kernell | With Kernell | Result |
| ------ | --------------- | ------------ | ------ |
| **Cost** | $0.25 | $0.03 | 💰 **88% cheaper** |
| **Latency** | 3.2s | 1.9s | ⚡ **40% faster** |
| **Quality** | Baseline | Same | ✅ **same quality** |

---

## 🧩 What You Get Out of the Box

* intelligent routing engine
* PolicyLite + Classifier-Pro
* DevLayer (built-in development)
* dashboard UI
* telemetry system
* benchmarking pipeline
* semantic cache
* CLI tooling
* version manager + OTA updates

---

## 🚀 What Makes Kernell Different

### 1. It’s an Environment, Not a Tool
Everything lives inside one system.

### 2. It Has a Native Economy
Most AI systems ignore value flow. Kernell tracks value, optimizes cost, and enables payments.

### 3. It Makes Decisions Automatically
You don’t choose models. Kernell does.

### 4. It Learns From Real Usage
Not static configuration — continuous improvement.

### 5. It Has a Human Interface
Dashboard = control layer. Not just monitoring.

---

## 🚀 Start Now

```bash
pip install kernell-os-sdk
kernell init --full
kernell dashboard
```
👉 See real value in under 2 minutes.

---

## ⚡ Final Statement

Kernell OS is not a wrapper. It is not a library. It is not a tool.

It is:
> **An operating environment for intelligent systems — with its own economy, protocol, and control layer.**
