​

Kody Concepts

Welcome to the Kody documentation. This guide provides an overview of key concepts within Kody, including Triggers, Agents, Tools, Kody Intelligence, and Integrations. These components work together to create a seamless experience for optimizing team workflows, enhancing productivity, and fostering collaboration.

​

Triggers

Triggers are the mechanisms that initiate interactions with Kody. They allow users to leverage Kody and its sub-agents in ways that fit seamlessly into their workflows. There are two primary types of triggers:

​

Conversation-Based Triggers

With conversation-based triggers, you can interact with Kody through a conversational interface, either using Kody’s web UI or through integrated communication platforms like Slack or Discord. This allows you to ask Kody questions, request specific actions, or seek insights in real time, making it a powerful assistant during daily team interactions.

​

AI Workflow Triggers (Automations)

AI Workflow Triggers are pre-configured automations that activate specific agents based on predefined conditions. You can customize these workflows to match your team’s needs, enabling agents to perform tasks automatically at the right time. For example, you could set an automation to have Kody review pull requests every day or generate weekly reports on team performance. These automations help reduce manual work and streamline repetitive processes.

​

Agents

Agents, sometimes referred to as sub-agents, are specialized modules within Kody, each designed to tackle specific tasks autonomously. Since Kody is the primary agent, these sub-agents function under its umbrella, handling unique workflows to address various needs in the development process.

Each agent is equipped with a dedicated set of instructions, workflows, and logic that allow it to perform distinct tasks. Here’s how they work:

• Autonomous Functionality: Agents operate independently, executing tasks without manual intervention once triggered.
• Task Specialization: Each agent is designed for a specific purpose, such as code review, documentation management, or performance monitoring, ensuring precise, high-quality outcomes.
• Customizable Workflows: Agents come with default workflows that can be customized, allowing you to tailor each agent to your team’s exact requirements.

Examples of agents include CodeReviewer, which automates the review process for pull requests, and TeamCoach, which monitors team performance metrics and offers insights for improvement.

​

Tools

Tools are the building blocks that agents rely on to complete their tasks. Each tool is a discrete, functional unit that represents a specific action or operation. Tools can include actions such as fetching data from a repository, updating documentation, or sending a notification.

​

Key Characteristics of Tools:

• Action-Oriented: Tools perform specific, isolated actions that agents can combine to complete more complex tasks.
• Reusable and Modular: Tools are designed to be reusable, allowing different agents to access them as needed.
• Integration-Ready: Tools are configured to work seamlessly with various integrations, giving agents access to external data sources and platforms.

Examples of tools include getPullRequest, which retrieves the latest pull requests from a source repository, and sendMessage, which posts updates to communication channels like Slack.

​

Kody Intelligence

Kody Intelligence is the central processing unit that powers Kody’s decision-making capabilities. This core intelligence layer analyzes data from multiple sources to provide agents with real-time insights and recommendations, enabling more informed, data-driven actions.

​

Kody Intelligence Features:

• Data Aggregation: Gathers information from multiple integrated systems to create a holistic view of team performance and project status.
• Real-Time Analysis: Processes data continuously to identify trends, detect issues early, and provide actionable insights.
• Insight Generation: Using advanced algorithms and machine learning models, Kody Intelligence surfaces recommendations to improve team efficiency, identify bottlenecks, and optimize workflows.

Kody Intelligence is what enables Kody and its agents to move beyond simple automation, offering a proactive and predictive layer that helps teams make better decisions.

​

Integrations

Integrations are the external platforms and tools that connect with Kody, providing the data and context required for agents to function effectively. By linking Kody to the tools your team already uses, such as project management, version control, and communication platforms, integrations enable a smooth flow of information that enhances the accuracy and usefulness of Kody’s outputs.

​

Integration Capabilities:

• Seamless Data Access: Integrations allow agents to retrieve and update data in real time, ensuring that Kody always has the most up-to-date information.
• Broad Compatibility: Kody integrates with a variety of platforms commonly used by engineering teams, including GitHub, Jira, Slack, Confluence, and more.
• Automated Workflows: Integrations allow agents to perform actions within external platforms automatically, such as creating issues in a project management tool or posting updates to a communication channel.

By leveraging integrations, Kody provides a unified workspace where all relevant tools and data sources converge, helping teams stay aligned and informed without needing to switch between multiple applications.