kanban-worker

Guide Kanban workers through lifecycle compliance and tenant-safe handoffs.

Updated May 5, 2026
One-click install
npx skills add https://github.com/Z43L/zeus-agent --skill kanban-worker-z43l
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: kanban-worker
Source: https://github.com/Z43L/zeus-agent/tree/main/skills/devops/kanban-worker
Command: npx skills add https://github.com/Z43L/zeus-agent --skill kanban-worker-z43l

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill prevents Kanban worker failures by documenting lifecycle expectations, correct workspace handling, isolation rules, and how to structure handoffs so orchestrators can interpret outcomes reliably.

Core Features & Use Cases

  • Workspace handling: Guides behavior across scratch, shared persistent directories, and Git worktrees so workers don’t corrupt state or lose progress.
  • Tenant isolation: Ensures persistent memory is namespaced to avoid cross-tenant data leaks.
  • Reliable handoffs: Defines how to call completion and blocking tools with well-shaped summaries and metadata for downstream parsing and dashboards.
  • Edge-case resilience: Explains fast-answered block reasons, heartbeat conventions, and retry diagnostics using prior run outcomes.

Quick Start

Load the kanban-worker Skill when dispatched by the Kanban dispatcher so your system prompt includes KANBAN_GUIDANCE and you can use kanban_complete, kanban_block, and heartbeats with correct workspace and tenant-safe behavior.

Frequently Asked Questions about kanban-worker

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I handle workspace state correctly in multi-agent kanban workflows?

Kanban multi-agent workflows require careful workspace handling across scratch, shared persistent directories, and Git worktrees to prevent state corruption. Workers must follow lifecycle expectations to ensure progress is not lost during task execution.

What is the best way to format handoff metadata for kanban task completion?

Kanban handoff metadata should be structured deterministically using kanban_complete and kanban_block outputs. This ensures orchestrators can reliably parse outcomes and dashboards can report progress accurately across multi-agent runs.

How do I avoid repeating failed paths in kanban retry handling?

Kanban retry handling requires using diagnostics from previous run outcomes to avoid repeating failed execution paths. Workers should examine prior outcomes before retrying to ensure they take a different approach on the next attempt.

Does kanban multi-agent execution support tenant isolation for persistent memory?

Kanban multi-agent execution supports tenant isolation by namespacing persistent memory to avoid cross-tenant data leaks. This ensures isolated agents working on different tenants do not access or overwrite each other's persistent data.

Why do kanban workers need to follow the orient, work, and block-complete lifecycle?

Kanban workers need lifecycle compliance because Hermes-style multi-agent runs require structured orient, work, and block-complete phases. This lifecycle ensures orchestrators interpret outcomes reliably and workers handle edge cases like fast-answered block reasons correctly.