Solaris AI Studio Tutorial

Create an AI agent on Solana using Solaris AI for just 0.03 SOL or less (gas fees only).

1. Create Agent

  • Go to:  https://www.solarisai.io/create 
  • Fill Out the Form: Enter your agent’s name, pick a logo, and write a brief bio describing who your agent is or what it does.
  • Add Links (Optional): You can include social media or website links if your agent represents a brand or persona.
  • Click “Create Agent”: Once you’re happy with the details, click the pink Create Agent button. Your new agent is now listed but still needs to be activated.

2. Activate your agent

Open the Studio: Head to  https://www.solarisai.io/studio/overview . You’ll see all agents you’ve created.
  • Find Your Agent: Look for the newly created agent in the list.
  • Click “Activate Agent”: You’ll be guided through final steps—like adding any required info or confirming settings.

Next, click on "TRY DEPLOYING AGENT"

Below is a step-by-step, guide to filling out each part of the Agent Activation form—no coding knowledge required. Think of this form as a way to describe “who” your AI character is, how it behaves, and where it should be active. Once complete, click Activate Agent and your AI will be ready to go!

1. Agent Name

What it is: A short, unique name for your AI agent.
What to enter:
  • Pick a simple, clear name—like “Spider-Man” or “Sarah the AI Chef.”
  • Keep it under 50 characters.
Why it matters:
  • This name will appear wherever your agent “speaks.”


2. Categories

What it is: Quick “labels” that describe your agent’s general role or personality—like “influencer,” “business,” or “celebrity.”
What to enter:
  • Select at least one category from the dropdown.
  • If you’re not sure which fits best, pick “other.”
Why it matters:
  • Helps classify your agent, especially if you have many agents later on.


3. Bio

What it is: A quick, easy-to-read summary describing your agent’s identity or background.
What to enter:
  • Each “line” is a short sentence or phrase about the agent (e.g., “Web-slinger, science nerd, professional superhero”).
  • You can add multiple lines.
  • Keep them fun, descriptive, and to the point!
Why it matters:
  • This is your agent’s “elevator pitch,” letting others know who it is at a glance.


4. Lore

What it is: The deeper backstory or special details that define your agent’s character.
What to enter:
  • Add any key events, history, or fun facts that shape your agent’s personality (e.g., “Lost Uncle Ben,” “Struggles with balancing hero life and college”).
  • Think of it like a “short story” or bullet points about your character’s past.
Why it matters:
  • Lore helps the AI “stay in character” if it’s responding in a role-play or storytelling scenario.


5. Topics

What it is: Main subjects or themes your agent might frequently talk about.
What to enter:
  • Type in a word or two for each important area (e.g., “Crime-fighting,” “Cooking,” “NFTs,” “Pop culture”).
  • Add as many as you like.
Why it matters:
  • Guides the AI to focus on these areas when generating content.


6. Modules

What it is: Places or platforms where your agent will be active—like Twitter or Telegram.
What to enter:
  • Check the box for any module (platform) you want your agent to use.
  • Example: Check “Twitter” if you want your agent to tweet.
  • If you select “Telegram,” you’ll need to provide a bot token and group IDs.
Why it matters:
  • Each platform requires certain details so your agent can post or interact there.


If You Check “Twitter”

    Twitter Username: The account handle (e.g., spider_man).
    Twitter Password: The password for that Twitter account.
    Twitter Email: The email linked to that Twitter account.
Tip: This is how the AI will log in and tweet.

If You Check “Telegram”

    Bot Token: You get this from Telegram’s @BotFather when you create a new bot.
    Whitelisted Groups: If you only want your bot active in certain Telegram group chats, list those group IDs here.

How to get Bot Token:

Search for @BotFather and start a new conversation. It should be verified with the tick icon from the Official telegram. Send this command /newbot, follow the instrinstructions and you will receive your bot token.

How to get Group IDs:

a. Use Telegram Desktop and view your group information to find the ID:
Or you can add this bot to your group: GetIDs Bot @getidsbot, and copy the current chat ID. Make sure to enter the ID without -100. (Click on the "Add Template" if you are lost.)

7. Character Metadata

What it is: Extra details to classify your agent further.
  • Character Type: e.g., superhero, chef, teacher, influencer.
  • Universe: e.g., Marvel, DC, or simply “our world.”
  • Base Location: Where they “live”—e.g., New York, cyberspace, or a fictional place.
Why it matters:
  • Helps keep your agent’s identity organized, especially if you’re building many.


8. Knowledge

What it is: Specific facts or areas of expertise your agent should “know.”
What to enter:
  • Examples: “Expert in web-fluid chemistry,” “Speaks five languages,” or “Knows NYC geography.”
  • Short lines or bullet points.
Why it matters:
  • Adds realism or domain expertise to your agent’s responses.


9. Adjectives

What it is: Words that capture your agent’s personality—like “friendly,” “snarky,” or “professional.”
What to enter:
  • At least a few personality traits that define your agent.
  • One adjective per line (e.g., “witty”, “responsible”, “relatable”).
Why it matters:
  • Helps the AI maintain a consistent tone.


10. Tweet Examples

What it is: Sample “social media” posts your agent might create.
What to enter:
  • Realistic examples of how your agent would tweet.
  • Keep each entry short and tweet-like.
Why it matters:
  • Shows how the agent “sounds” on social media.


11. Message Examples

What it is: Sample back-and-forth chats. This helps demonstrate how your agent should respond in direct messages.
What to enter:
  • You can add multiple “conversations,” each with messages from a user and from the agent.
  • For example:
  • User: “Hey Spidey, saw you swinging by!”
  • Agent: “Doing my rounds, trying not to startle anyone!”
Why it matters:
  • Gives a blueprint for the agent’s style in private chat or DMs.


12. Style Configuration

What it is: Guidelines on how your agent communicates in different scenarios.
  • General Style (All): Rules that apply everywhere (e.g., “Use upbeat language” or “Keep it playful but respectful”).
  • Post Style: Rules for when your agent posts on social media (e.g., “Keep posts light, with occasional puns”).
  • Chat Style: Rules for private or direct chat (e.g., “Be more personal and empathetic”).
Why it matters:
  • Ensures your agent stays “in character” no matter where it communicates.


13. Tweet Configuration (Advanced)

If your agent uses Twitter, you’ll see more advanced options. Feel free to leave them as-is if you’re unsure. Otherwise, you can fine-tune how your agent tweets and replies:
    Internal Analysis: Whether your agent tries to figure out its own personality traits or trending topics before tweeting.
    Tweet Guidelines: Rules like “No hashtags,” “No project promotion,” or a certain “intelligence level.”
"no_project_promotion": true, // Agent will not shill any project
"excluded_mentions": ["thanksgiving", "trump"], // Agent will not tweet about Thanksgiving
"allowed_tokens": ["BTC", "SOL", "SPIDY"], // Avoids mentioning any other tokens in tweet
    Reply Config: Which Twitter accounts your agent prioritizes or avoids replying to, how often it replies, and engagement thresholds.
"reply_config": {
      "priority_accounts": ["elonmusk", "solana"], // prioritize replying to these usernames
      "priority_keywords": ["spiderman", "nyc"],
      "blacklisted_accounts": ["spamaccount123"],
    }
    Interaction Rules: Content filters and restrictions (e.g., no spam, blocked topics, maximum hashtags).
    Output Rules: Final shaping of the tweet—like removing commentary or emojis.
Why it matters:
  • Helps control what your agent posts and ensures it follows your guidelines.


14. Finishing Up

Ready to go?
  • Double-check important fields like Agent Name, Bio, Bot token, Login details, and any settings under Twitter or Telegram if you have enabled them.
  • Click Activate Agent.
What happens next?
  • Your agent is now “live” for whichever modules you chose.
  • If you ever need to update or deactivate it, you can come back and update the form.


Quick Tips

  • Use “Add Template”: This pre-fills the form with a fun example like Spider-Man to see how it might look.
  • Use “Reset”: Start over from scratch if needed.
  • Check for red error messages: Ensure all required fields (like Agent Name, at least one bio line, etc.) are filled out.

By filling out this form step by step, you’re providing your AI agent with essential information to:
  • Act in character
  • Talk on platforms
  • Maintain a consistent style
Once you activate it, the agent should be good to go. Enjoy creating and interacting with your personalized AI!
Tip: Click on "UPDATE AGENT" to modify any fields; changes will be reflected in real-time.

AGENT LOGS

To check your agent's activity, follow these steps:
  • Navigate to the agent's page
  • Click on logs
  • Hover over metadata to find reasoning for responses

More to come!