Armor Wallet Docs
  • Introduction
  • 💚Armor BASICS
    • Vision
    • Token
    • FAQ
    • Codex NFT
    • Platform Fees
    • Supported Chains
    • Supported Protocols
    • Core Functions
      • Wallet Management
      • Security
  • 🐸Armor Agents
    • Creating an Agent
  • Agent Customization
  • Agent Blueprints
  • ✳️Armor SDK
    • SDK Introduction
  • Armor MCP
  • Early Access Armor API key
  • Tutorials
    • Installing the Armor MCP
    • Troubleshooting UV and UVX
    • n8n
    • Claude
    • OpenAI
    • Langraph
    • Cline
    • Cursor
    • ElizeOS
    • Pydantic AI
  • 🟩AI & Tech
    • Architecture
    • Multi-Agent System
    • Agent Ecosystem
      • Trading Agent
      • Research Agent
      • Portfolio Agent
    • Game Connect
    • Bug Bounty
  • 🟢Company
    • Brand Kit
    • Work with us
Powered by GitBook
On this page
  • Tone of voice
  • Behavior
  • Spcialization
  • Strategies
  • Functions
  • Skys the limit

Agent Customization

This is an overview of the some of the possibilities you can explore with custom agents.

Armor Wallet give all users the ability to create complete custom AI Agents. This customization is unlike anything you have experiences before. The customization include features, functionality, behaviors, strategies and more. Because this is customizing the actual AI, the sky is the limit as to the possibilities.

Be aware with customization of agents you have the power to confuse the AI and cause it to provide inaccurate or faulty results. Be careful when writing your Agent Blueprint and build off of best practises.

To start with this guide will walk you through some of the basic ways to customize your agents.

Tone of voice

Tone of voice defines how your agent sounds when it talks to you. It has a huge impact on how comfortable and usable your experience feels. Do you want your agent to be professional and concise, like a financial advisor? Or do you prefer a more casual, friendly vibe that explains things in plain English? You can even go full meme-mode and make it speak like a crypto degen or use emojis and slang. The tone of voice should match your personality and your comfort level with crypto. This helps ensure the AI feels like a natural extension of how you think and talk.

Behavior

Behavior is about how your agent acts especially when making decisions or responding to risk. It includes things like when it should stop and ask you for confirmation, how cautious or aggressive it should be when evaluating trades, and whether it should act autonomously or always double-check with you first. You can also define ethical guardrails: for example, telling the agent never to recommend highly volatile assets unless asked, or always calculate risk-to-reward before suggesting a move. This helps prevent unwanted surprises and ensures the AI operates in a way you trust.

Spcialization

Specialization determines what your agent is good at. Think of it as its role or niche. Some agents might specialize in portfolio rebalancing and long-term wealth building, while others might be experts in finding short-term trading opportunities based on momentum. You can also define research-focused agents, meme coin hunters, or data analysts that help you understand token fundamentals. Giving your agent a specific area of expertise helps it stay focused and ensures it gives the most relevant responses for your use case.

Strategies

Strategies are like pre-defined processes or playbooks, a complete sets of actions and filters that the agent can execute whenever you call on them. Instead of manually walking through every step each time you want to research or invest, you can give your agent a strategy that wraps all of that complexity into a single command. For example, you might define a strategy called “Find New Gems” that tells the agent to scan new tokens from a specific launchpad, filter for high liquidity, and analyze 1-hour candlestick trends to spot early upward momentum. Or you might define a “Pullback Entry” strategy that watches for tokens dropping 10–15% from a local high with strong support levels. These strategies are not just settings, they’re reusable methods for how your agent should think and act across specific investing scenarios. Once defined, you can reference them by name and apply them to different market situations, saving time and making your process more consistent and powerful..

Functions

Functions are the practical tools and commands your agent uses to do its job. These include getting wallet balances, showing open positions, fetching trending tokens, simulating trades, calculating profits and losses, and more. In your Blueprint, you can define shortcuts like typing “wallets” to see all your balances, or “trending” to fetch hot tokens based on 1-hour charts. By specifying these functional behaviors, you make the agent faster, more predictable, and easier to use, especially for frequent tasks you perform every day.

Skys the limit

The above are guides to help you better understand some of the possibilities of customization but by no means the definitive answer. You can explore and see what you can do with Agent Blueprints, if you find a novel use, please share it with us and we can showcases it.

PreviousCreating an AgentNextAgent Blueprints

Last updated 3 days ago