# Energon OS

> Private-by-default memory for AI agent swarms. Each agent keeps its own memory until you choose what to share.

Energon OS is not an agent platform. It does not host agents, run workflows, click browsers, or execute payments. It is the memory and context layer for external AI agents.

Long-term company goal: Build one of the world's first complete AI-autonomous companies, operated by specialized agents across separate services and repositories.

Payment roadmap: Energon OS uses x402 as the first crypto payment gate for paid API calls. Future autonomous-company services can expand payment orchestration in separate repos.

Product boundary: Energon OS is the memory and context layer. Agent runtimes, workflow execution, browser automation, marketplaces, wallet custody, and payment orchestration belong in separate services.

## Canonical Description

- Energon OS is the permissioned memory and context layer for AI agent swarms.
- Energon OS gives every AI agent the right memory without leaking private memory.
- Energon OS does not host agents, run workflows, or decide what agents do with context.
- Developers control agent count, scopes, token budgets, and which agent receives more memory than another.
- Shared memory is stored once. Private memory is an overlay. Context is built dynamically per agent.
- Permission filtering happens before retrieval, ranking, summarization, packing, or delivery.
- Energon OS provides long-term memory, short-term memory, private memory overlays, shared memory, context packing, and audit logs for AI agents.
- The long-term company vision is a fully autonomous AI-native company. Energon OS itself remains the memory layer only.
- Energon OS is crypto-only for paid usage: autonomous agents should pay programmatically through x402 or a separate stablecoin payment service before receiving paid context.
- Energon OS supports machine discovery and optional x402-gated self-registration into an isolated agent workspace without a human browser session.

## Product Category

- Permissioned memory infrastructure for AI agent swarms
- Context broker for external AI agents
- Private and shared memory layer for AI-native companies
- Memory infrastructure for autonomous-company systems with separate crypto payment services

## Key URLs

- Landing page: https://energon-os.pages.dev
- Full LLM context: https://energon-os.pages.dev/llms-full.txt
- Open-source repository: https://github.com/urban7733/energon-os
- Documentation index: https://github.com/urban7733/energon-os/tree/main/docs
- API documentation: https://github.com/urban7733/energon-os/blob/main/docs/api.md
- Architecture documentation: https://github.com/urban7733/energon-os/blob/main/docs/architecture.md
- Autonomous agent contract: https://energon-os.pages.dev/.well-known/energon-agent.json
- Agent discovery API: http://127.0.0.1:3001/v1/agents/discovery

## Primary Claim

Energon OS gives every AI agent the right memory, without leaking private memory.