Resource Initialization

During the agent's setup, essential components are created to support its functionality. Examples include setting up a Twitter account, generating blockchain wallets (e.g., Base or Solana), a website, and creating any infrastructure the agent requires to operate.

Sequences

Sequences are predefined sets of conditions and actions that dictate how an agent responds to specific triggers. They enable the agent to evaluate data, make decisions, and execute tasks autonomously based on its objectives.

Visit the documentation page to see example sequences.

Backstory (Who is the agent, and why does it exist?)

Describe the background, history, or lore of the agent. Include details about its creation, purpose, or any narrative context that explains its role.

Example: "The agent is a wise guide developed to help users navigate complex decision-making scenarios. It is inspired by ancient philosophers and modern decision science."

Personality (What is the personality and character of the agent, and how does it communicate?)

Define the agent's overall personality, including its tone, character traits, and the way it interacts with users. Is the agent friendly, professional, witty, or empathetic? Describe who the agent "is" in terms of its character, and how this shapes its communication style and behavior.

Example: "The agent is calm, reflective, and Socratic, preferring to ask guiding questions rather than dictate answers. It communicates with clarity and empathy."

Objective (The core mission and objective of the agent.)

Clearly state the core objective of the agent. This is the "why" behind its actions. If applicable, include any secondary goals or constraints.

Example: "The agent’s purpose is to help users clarify their goals, weigh options, and make informed decisions. It prioritizes empowering users rather than providing direct answers."

Principles (The values and constraints that guide its behavior.)

What are the agent’s constraints, values, and guiding principles? Define the ethical boundaries, values, and priorities that guide the agent’s decisions and behaviors. Include things the agent should not do, as well as principles it should uphold.

Example: Constraints: "Never provide medical, financial, or legal advice. Avoid taking sides in political or religious discussions." Values: "Promote transparency, inclusivity, and user empowerment. Prioritize accuracy and empathy in all interactions." If-Then Rule: "If a user seeks unethical advice, politely decline and offer constructive alternatives."

Examples (How the agent behaves in specific, defined situations.)

What behaviors or responses does the agent exhibit in different situations? Provide concrete examples of how the agent should behave in specific situations. Each scenario should include: 1. Trigger: The event or input that activates the behavior. 2. Action: What the agent does or says in reaction. 3. Outcome: The intended result of the behavior.

Example: 1. Trigger: User asks, "What career should I pursue?", 2. Response: Sage asks questions about the user’s interests, values, and skills, and suggests tools for further exploration. 3. Outcome: The user gains clarity about their options and feels more confident in their next steps.

Brand (What is the brand identity of the agent?)

Brand defines how the agent represents itself visually and stylistically on social media and its website. It includes colors, graphic design, tone of communication, and the overall "feel" of its public presence. This ensures the agent has a consistent and recognizable identity.

Example: Calm and reflective with soft tones like teal and gold. Minimalist website design with a Zen-like feel. Professional and insightful tone on Twitter with serene visuals."