An agent is an AI system that reasons through problems and executes tasks autonomously. Agents are configured by specifying models, tools, instructions, and reasoning approach—providing everything needed to accomplish tasks from simple workflows to complex, open-ended objectives.

Agent primitives

Agents are built from two core primitives:

Agent configuration

When creating an agent, you configure several key properties:

• Name – Identifier for the agent
• Instructions – System prompt that defines how the agent should generate its response (also called developer message)
• Model – Which LLM to use for generating responses (can be a base model or fine-tuned model)
• Tools – Which tools the agent can access
• Reasoning effort – Adjusts the performance and reasoning tradeoff, controlling how much reasoning the agent uses for planning

Reasoning modes

Agents support two reasoning modes optimized for different scenarios:

Speed-optimized – Streamlined reasoning for low-latency applications. Best for scenarios requiring quick responses with a limited set of tools.

Performance-optimized – Enhanced reasoning for complex scenarios. Essential for agents managing multiple tools or executing sophisticated workflows where accuracy outweighs response speed.

Agent status

Each agent has a status that indicates its deployment state:

• Active – Ready to serve requests
• Pending – Recently promoted, transitioning to active state
• Inactive – Not currently deployed, requires activation before serving requests

Creating agents

Agents can be created and managed through:

• SeekrFlow UI – No-code interface for creating, managing, and testing agents
• SeekrFlow API – Programmatic access for integration into applications
• SeekrFlow SDK – Python SDK for building complex reasoning systems