UNPKG

@tosin2013/kanbn

Version:

A CLI Kanban board with AI-powered task management features

59 lines (52 loc) 3.15 kB
Usage: kanbn chat [options] kanbn c [options] Description: Start an AI-powered chat with a project management assistant that understands your Kanbn board. The assistant maintains context across interactions, preserves chat history, and ensures board state consistency. It can help explain your project, analyze tasks, manage relationships, and provide guidance on project management. Options: -m, --message Specify a one-off message to send to the assistant -k, --api-key Specify the OpenRouter API key (overrides environment variable) --model, --mod Specify the OpenRouter model to use (overrides environment variable) --with-refs Include task references in the context --include-refs Same as --with-refs --references Same as --with-refs --with-integrations Include all available integrations in the context --integrations Same as --with-integrations --use-integrations Same as --with-integrations --integration <name> Include a specific integration by name --use-integration Same as --integration --integrate Same as --integration -h, --help Show help Examples: kanbn chat kanbn c kanbn chat --message "Create a new task called Project Setup" kanbn chat --message "Add three subtasks to that task" kanbn chat --message "Move the first subtask to In Progress" kanbn chat --message "Summarize my current project status" kanbn chat --api-key "your-api-key" --message "What should I work on next?" kanbn chat --model "openai/gpt-4o" --message "Analyze my project and suggest improvements" # Integration examples kanbn chat --with-integrations --message "What stats are used in D&D 5E?" kanbn chat --integration game-systems --message "Tell me about character stats in D&D" kanbn chat --integration api-docs --integration database-schema --message "How do I query users?" Notes: - Requires an OpenRouter API key set as OPENROUTER_API_KEY environment variable or via --api-key option - Can specify a model via OPENROUTER_MODEL environment variable or --model option (default: google/gemma-3-4b-it:free) - Maintains context across chat sessions for natural conversations - Preserves board state consistency during operations - Handles task relationships and dependencies - Provides error recovery and fallback behaviors - Logs AI interactions for tracking and analysis - Interactive mode allows for continuous conversation - The assistant has access to your project context including tasks, columns, and statistics Integration Features: - Uses RAG (Retrieval-Augmented Generation) to enhance AI responses with domain knowledge - Integrations are stored as markdown files in the .kanbn/integrations directory - Manage integrations with the 'kanbn integrations' command - Use --with-integrations to include all available integrations in the context - Use --integration <name> to include specific integrations - Supports local Ollama models for embeddings when available - Falls back to simple keyword matching when embeddings are not available