Book
AI Native Platform
An operator's walkthrough of the open platform that runs AI agents, the software helpers that do the work for you, on your own computer. Twenty one chapters in five parts, from the shop floor to every API call. Free to read online.
- 21 chapters
- 5 parts
- ~115,000 words
- Length
- 21 chapters
- Parts
- 5 parts
- Size
- About 115,000 words
- Price
- Free to read
AI Native Platform
What's inside
Part 1 · Orientation
- 1
The Shop Floor
A first look at the whole platform, and how a plan turns into the real work the agents do.
- 2
The Five Station Tour
A quick tour of the five main parts, so the words in the rest of the book make sense.
Part 2 · Running the Factory
- 3
Task Execution
How a single job, written on a card, gets picked up and run by an agent.
- 4
Workflows and Blueprints
How to chain many jobs into one flow that runs step by step.
- 5
The World Model
The shared memory of files and tables that every agent can read and write.
- 6
Schedules and the Heartbeat
How to set work to start on a clock, so the platform keeps running on its own.
- 7
Chat as a Posture
How a plain chat with you can hand the agents new work at any time.
Part 3 · The Intelligence Layer
- 8
Profiles
How each agent gets its own role, voice, and rules to follow.
- 9
Memory and Handoffs
How an agent keeps what it learned, and passes a job cleanly to the next one.
- 10
Monitoring and Cost
How to watch every run and see what it costs before you scale up.
Part 4 · The API in Depth
- 11
The Task API
The web calls (an API, the way one program asks another to do something) that create and run one task from your own code.
- 12
The Workflow API
The calls that build and start a flow of many steps.
- 13
The Schedule API
The calls that set work to run on a timer.
- 14
The Trigger API
The calls that start work when something changes, like a new file or message.
- 15
The Tables API
The calls that read and write the shared tables of data.
- 16
The Document API
The calls that hand files to an agent so it can use them in its work.
- 17
The Conversation API
The calls that send and stream a back and forth chat with an agent.
- 18
The Profile API
The calls that set up and change an agent's role and rules.
- 19
The Memory, Handoff, and Notification APIs
The calls that save memory, pass work along, and send an alert when something needs you.
Part 5 · The Factory as a Whole
- 20
The Factory, Put Back Together
All the pieces working as one, shown as a small business that runs itself.
- 21
The Endpoint Reference
A full list of every API call across the nine parts of the platform, for quick lookup.
AI Native Platform is the operator’s manual for the open platform that runs AI agents on your own computer. Agents are software helpers you hand a job to, and they carry it out for you. This book walks you across the whole platform, one piece at a time, with real code behind every step. The platform is free to install, free to use, and free to fork.
What you will learn
You start on the shop floor, where a plan turns into a day’s work, then learn how a single task runs, how to chain tasks into a flow, and how to set that flow on a schedule so it runs on its own. From there you give each agent its own role and memory, watch every run and its cost, and finally call every part of the platform from your own code through its full API (the set of web calls one program uses to ask another to do something). The last part puts the whole factory back together as a small business that runs itself.
Who it is for
Builders who want to run AI agents on hardware they own, not rent, and wire them into their own tools. You do not need a cluster or a big budget. You can read the whole thing free online. If you want a copy to keep and read offline, the PDF and EPUB bundle is yours for a one time price.
Get the full book
PDF and EPUB, yours to keep.
$40one time
Not happy in the first 14 days? Email us and we refund you, no questions asked.
Keep exploring
AI Native Business
A book on running a business with AI agents (software helpers that do the work for you). Fourteen short chapters in four parts take you from the first idea to a working system. About a two hour read, open to all.
One agent, three faces
The same AI agent should meet a developer in the terminal, a manager in a dashboard, and a phone user by voice. One brain, many doors.
A spec at breakfast, an app by lunch
With Kiro I turned three sentences into a 24,000-line app, about four hours of my time. The spec did the heavy lifting, and speed became a skill.