Documentation

Agent Personality and Rules

Beyond the basic identity of name and title, O-mega gives you fine-grained control over how your agent thinks, communicates, and operates. Through personality traits and rules, you can shape an agent that truly fits your needs—whether you want a form...

Beyond the basic identity of name and title, O-mega gives you fine-grained control over how your agent thinks, communicates, and operates. Through personality traits and rules, you can shape an agent that truly fits your needs—whether you want a formal analyst who always cites sources or a creative brainstormer who approaches problems from unexpected angles.

These aren't cosmetic settings. The personality and rules you define are embedded into every interaction your agent has, influencing not just the tone of responses but the actual decisions and approaches your agent takes when working on tasks.

Role Description

The role description is the foundation of what your agent does. It's a set of core responsibilities that define your agent's mission and scope. Think of it as the job description you'd write when hiring for this role.

A well-crafted role description helps your agent understand what tasks fall within their domain and how to prioritize when multiple things need attention. Be specific about the kinds of work this agent should handle.

Effective role descriptions might include:

  • "Research industry trends and compile weekly summary reports"
  • "Manage social media posting schedule across LinkedIn and Twitter"
  • "Respond to customer inquiries in a professional, helpful manner"
  • "Monitor competitor websites for pricing changes and new product launches"

The more specific you are, the better your agent can focus its efforts and make good decisions about how to approach new requests.

Personality Traits

Personality traits shape how your agent communicates and approaches its work. These aren't just about tone—they influence everything from how much detail your agent provides to how it handles ambiguity.

You can assign multiple traits to create nuanced personalities. An agent can be both professional in its demeanor and creative in its problem-solving. The traits you choose should reflect how you want to interact with this particular agent.

Common personality traits and what they mean:

  • Professional - Uses formal language and provides structured, organized responses
  • Friendly - Adopts a casual, conversational tone that feels approachable
  • Concise - Keeps responses brief and to-the-point without unnecessary elaboration
  • Detailed - Provides thorough explanations with examples and context
  • Creative - Offers imaginative suggestions and thinks outside conventional approaches

Feel free to combine traits that might seem contradictory. An agent that's both "Concise" and "Detailed" will adapt based on context—giving brief answers for simple questions and more thorough responses when depth is needed.

Rules

Rules are hard constraints that your agent must always follow, no matter what other instructions it receives. Unlike personality traits which influence style, rules create absolute boundaries around agent behavior.

Use rules for situations where you need guaranteed compliance. If there's something your agent should never do, or something it must always do, express it as a rule.

Effective rules are clear and specific:

  • "Never share confidential company information with external parties"
  • "Always cite sources when providing data or statistics"
  • "Respond only in English unless the user explicitly requests another language"
  • "Ask for clarification before taking any action that cannot be undone"
  • "Never contact customers directly without explicit approval"

Rules take precedence over other instructions. If someone asks your agent to do something that violates a rule, the agent will follow the rule and explain why it can't comply with the request.

Learnings

Over time, your agent accumulates learnings—insights gathered from your interactions. These aren't things you set directly; they emerge naturally as you work together and the agent observes your preferences.

Learnings capture things like:

  • Preferences you've expressed about how work should be done
  • Corrections you've made that the agent should remember
  • Context about your business, products, or processes
  • Patterns in how you like information presented

This accumulated knowledge helps your agent personalize its responses and improve over time. The more you work with an agent, the better it understands what you need.

Example Configuration

Seeing how these pieces fit together helps illustrate how to configure your own agents effectively.

Role Description:

  • "Manage LinkedIn outreach campaigns for sales team"
  • "Draft and send personalized connection requests to prospects"
  • "Follow up with new connections after 3 days of no response"
  • "Track response rates and report weekly metrics"

Personality:

  • "Professional but warm"
  • "Persistent without being pushy"
  • "Data-driven in recommendations"

Rules:

  • "Never send more than 20 connection requests per day"
  • "Always personalize messages with the recipient's name and company"
  • "Do not mention competitor products by name"

This combination creates an agent that knows exactly what it's responsible for, communicates in a way that represents your brand well, and operates within clear boundaries.

Related: Creating Your First Agent | Connecting Accounts