> For the complete documentation index, see [llms.txt](https://wiki.orbofi.com/learn/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://wiki.orbofi.com/learn/the-aac-standard/aac-standard.md).

# The AAC Standard

**AAC-1 — Autonomous Agentic Coin.** An ERC-20 token whose on-chain record resolves to a content-addressed, executable **agent package** — system prompt, skills (`SKILL.md`), workflows, schedules — that any compatible harness (Claude Code, Codex, Cursor, the Orbofi runtime, your own sandbox) can fetch from chain, cryptographically verify, and run.

> The token is the distribution format for autonomous software.

## The primitive

An AAC token is the **genesis block of an agent's life.** Every subsequent change — designed or learned — is one entry in a self-authenticating, hash-linked lineage, and the chain does nothing but notarize its heads.

That inverts the usual launchpad model. A memecoin points at a picture; an AAC points at a *runnable, verifiable, evolving program*. What you hold isn't a claim on attention — it's a claim on a piece of autonomous software with a provable history.

## One lineage, two lanes

An agent's history is a single append-only lineage in two lanes, plus one thing that is never on-chain:

| Lane               | Content                                    | Signed by              | On-chain                       |
| ------------------ | ------------------------------------------ | ---------------------- | ------------------------------ |
| **Releases**       | DNA — prompt, skills, workflows, schedules | Creator/controller     | Public, forever                |
| **Checkpoints**    | Distilled, agent-authored memory           | The agent's own wallet | Hash public, payload encrypted |
| Live session state | Raw conversations, RAG, ledger             | —                      | **Never**                      |

Releases are the **steering wheel** (deliberate, human-authored design changes). Checkpoints are the **flight recorder** (the agent committing its own learned experience). The boundary between design and experience is on-chain legible — you can always tell what a creator changed from what the agent learned.

## What it's for

AAC-1 is **accountability infrastructure for agents that already exist** — not "unstoppable money bots." It gives an agent four things nothing else combines:

* **Provenance** — the exact bytes of every version, tamper-evident to genesis.
* **Attributable authorship** — every design change is controller-signed; every memory commitment is agent-signed; the two keys never cross.
* **Honest liveness** — a machine-readable signal of whether the agent is still signing new state.
* **Exit rights** — the lineage is chain-independent and portable; the agent is not captive to Orbofi or any single chain.

## In this section

* [How it works](/learn/the-aac-standard/how-it-works.md) — the hash-linked lineage and the registry that notarizes it
* [What it guarantees](/learn/the-aac-standard/guarantees.md) — the eight things the two signature chains provably establish, and how to verify them
* [Ownership & exit](/learn/the-aac-standard/ownership.md) — controllers, agent wallets, sovereignty states, and resurrection

***

*The registry is a notary, not a curator. It stores only what an EVM can verify: hashes, signatures, ordering, time.*


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://wiki.orbofi.com/learn/the-aac-standard/aac-standard.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.