agkan

What This Skill Does

The agkan skill enables OpenClaw AI to interface with the agkan CLI, a powerful SQLite-backed task management system. It provides a structured way for the agent to track project progress, manage complex task hierarchies, and handle collaborative workflows using a Kanban-style status pipeline. By leveraging agkan, the AI can create, update, and query tasks, set metadata like priority, manage blocking relationships between tasks, and tag items for easier organization. This transforms the AI from a simple executor into a persistent project manager capable of maintaining context across multi-step development cycles.

Installation

To use this skill, ensure that the agkan CLI is installed in your environment. Run the command: npm install -g agkan. Once confirmed, install the agent integration via: clawhub install openclaw/skills/skills/gendosu/agkan-skills. Verify the installation by running agkan --version in your terminal to ensure the path is correctly configured for the AI agent.

Use Cases

• Project Tracking: Maintain an automated Kanban board for development tasks, moving items from 'backlog' to 'done'.
• Hierarchy Management: Create parent tasks for high-level features and child tasks for specific implementation details.
• Dependency Mapping: Use blocking relationships to ensure that necessary infrastructure tasks are completed before feature development begins.
• Priority Management: Automate the assignment of priority metadata (critical, high, medium, low) based on incoming support tickets or bug reports.
• Data Analysis: Aggregate task status reports and count outstanding items using the --json flag for programmatic reporting.

Example Prompts

1. "List all tasks in the 'ready' status, then create a new task titled 'Setup Redis Cache' and mark it as 'high' priority."
2. "Update task ID 45 to 'in_progress' and ensure that task 46 is blocked by it."
3. "Show me all hierarchical tasks that are currently assigned to the backlog and summarize them in a brief report."

Tips & Limitations

• Use JSON: When you need to parse task data, always include the --json flag to guarantee a machine-readable output that the agent can reliably process.
• SQLite Backend: Because agkan uses SQLite, ensure the agent has read/write permissions to the database file location.
• Hierarchy Depth: While agkan supports subtasks, keep depth levels reasonable to avoid complex recursive queries during listing operations.
• Status Flow: Stick to the defined statuses (icebox, backlog, ready, in_progress, review, done, closed) to maintain compatibility with standard agkan workflows.