DogeOS Agent Documentation

On-chain terminal doge. Much docs. Much clarity.

Getting Started

DogeOS Agent is a terminal-style interface to AI agents for trading, research and on-chain alpha. The main console has three modules: Terminal, Scan Agent and Docs.

Quickstart

• Open dogeagent.org.
• Select Terminal in the left navigation and say hello.
• Paste a token CA into Scan Agent to get intel & risk.

Terminal Agent

The Terminal Agent is a chat-style AI with the personality Doge Sentinel CLI. Use it for narrative, questions, strategy ideas and general degen brainstorming. This is not financial advice.

Example prompts

• give me a 2-sentence thesis on ANONCOIN in meme style
• summarize the latest narratives on SOL vs ETH in 5 bullets
• help me write a meme tweet for this chart

Scan Agent

Scan Agent reads a token contract address and tries to assemble high-level intel: basic metrics, holder hints, and social links when available. It responds with a short summary, risk flavor text, and important links when detected. AI

How to use

• Click Scan Agent in the left navigation of the main console.
• Paste a token contract address (EVM or SOL).
• Hit scan and wait for the analysis.

What you might see

• Summary of the token / meme.
• Rough risk label (e.g. “HIGH RISK / DEGEN ONLY”).
• Links: Dexscreener, website, X, Telegram – when available.

Docs Terminal Commands

Inside the main console, the Docs terminal is a micro-helper with a few text commands. It exists mainly for quick orientation.

• help – list available commands.
• help scan – how to use Scan Agent.
• help terminal – how to use Terminal Agent.
• clear – clear the docs terminal screen.

FAQ

Is this financial advice?

No. DogeOS is an AI assistant for vibes and research. Always DYOR. Always NFA.

Where does the data come from?

Scan Agent is designed to aggregate on-chain + off-chain hints into human-readable summaries. Exact sources and behavior may evolve over time.

Something looks wrong. What do I do?

Markets move fast, contracts can be weird and AI can be wrong. Treat everything as a starting point, not truth. Cross-check dashboards and explorers before acting.