Welcome to the developer documentation for the Cognotik Framework. Cognotik is a sophisticated AI-powered application framework designed for complex task planning, code execution, and orchestration.

This repository contains detailed guides ranging from high-level architectural strategies to specific implementation details for UI servlets and code interpreters.

Understanding the foundation of the system, including session management, storage, and the design philosophy behind the framework's extensibility.

• Platform Architecture
  • Overview of the layered architecture (Service, Interface, Implementation).
  • Session management (Global vs. User sessions).
  • Data storage, authentication, and cloud integration.
• Strategic Extensibility
  • Must Read: Explains the "Multiplicative Scaling" philosophy.
  • Deep dive into the Strategy Patterns used for Models, Providers, and Tasks.
  • Historical context on interoperability and AI-assisted development.
• User Settings Subsystem
  • Managing API credentials and local tool configurations.
  • Persistence, security masking, and JSON migration logic.
• Generative Use Cases
  • A taxonomy of generative AI patterns (0→1 Creation, 1→1 Transformation, Synthesis, etc.).

How the AI "thinks," assumes personas, and structures its reasoning.

• Actor Types
  • Guide to the base AI workers: SimpleActor (Text), ParsedActor (Structured Data), CodingActor (Code Gen), and ImageActor.
• Cognitive Modes
  • User guide to the planning engines: Conversational, Waterfall, Adaptive Planning, and Hierarchical Planning.
• Cognitive Schemas
  • Defining AI "Mindsets" or personas (e.g., Project Manager, Scientific Method, Agile Developer, Critical Auditor).

The core engine where AI logic meets actionable execution.

• Task Planning Overview
  • Start Here: Introduction to the Task Planning Framework.
  • Concepts of Tasks, Orchestration, and how they interact with Cognitive Modes.
• Task Library & Reference
  • Detailed documentation of available tools: RunCodeTask, FileModificationTask, CrawlerAgentTask, VectorSearchTask, and more.
• Launch API & Configuration
  • How to configure and launch sessions via the Web UI, IntelliJ Plugin, or HTTP API.
  • Details on OrchestrationConfig and model selection.

The subsystems responsible for running code and applying AI-generated changes to source files.

• Interpreter Subsystem
  • Guide to the Interpreter interface.
  • Implementations for Kotlin, Groovy, and Shell execution.
  • Security and output interception.
• Patch Processors
  • Deep dive into the fuzzy matching algorithms used to apply AI code edits.
  • Explanation of different processors (CStyle, Python, Markdown, Strict vs. Lenient).

Documentation for the web-based components of the framework.

• Servlets Guide
  • Overview of all HTTP endpoints (Auth, Session Management, File Serving, Proxying).
• Server-Driven UI
  • Guide to the SessionTask UI system.
  • How to build interactive, streaming UIs using Kotlin (Tabs, Buttons, Markdown rendering).
• Application API
  • Detailed reference for HTTP endpoints (Session management, File system, Usage).
• WebSocket Protocol
  • Technical guide to the real-time communication protocol.
  • Message replay, synchronization, and data transport.

Guidelines for extending the framework and maintaining quality.

• Task Implementation Standards
  • Critical for Contributors: Rules for creating new TaskType entities.
  • Safety protocols (Human-in-the-loop), UI feedback requirements, and configuration standards.
• Product Page Guidelines
  • Design and content standards for creating user-facing "Product Pages" for specific tasks.

1. New Users: Start with Task Planning Overview and * Cognitive Modes* to understand what the system does.
2. Architects: Read Strategic Extensibility and Platform Architecture to understand the design philosophy.
3. Task Developers: Read Actor Types, Task Library, and * *Task Implementation Standards**.
4. UI/Web Developers: Focus on Servlets and Server-Driven UI.