Secretary

Orchestrate AI sub-agents, worktrees, and daily context for Claude Code organizations.

Updated Dec 29, 2025
One-click install
npx skills add https://github.com/shabaraba/shabaraba-cc-plugins --skill secretary
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Secretary
Source: https://github.com/shabaraba/shabaraba-cc-plugins/tree/main/packages/claude-org/skills/secretary
Command: npx skills add https://github.com/shabaraba/shabaraba-cc-plugins --skill secretary

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This skill coordinates the Claude Code organization by managing AI sub-agents, worktrees, and daily context to keep operations flowing smoothly.

Core Features & Use Cases

  • Async agent orchestration and lifecycle tracking across multiple teams (product/engineering/QA/marketing)
  • Worktree creation, isolation, and branch management for concurrent development
  • Daily logs, context propagation, and handoff document generation for seamless handoffs

Quick Start

To start using Secretary, initialize a new worktree and context, then launch a task. bash ${CLAUDE_PLUGIN_ROOT}/scripts/work-manager.sh create-worktree "<branch>" bash ${CLAUDE_PLUGIN_ROOT}/scripts/work-manager.sh init-context "<branch>" bash ${CLAUDE_PLUGIN_ROOT}/scripts/work-manager.sh add-task ...

Frequently Asked Questions about Secretary

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

FAQPage Schema
How do I orchestrate AI sub-agents across multiple teams like product and engineering?

AI sub-agents are orchestrated by coordinating task launches, worktrees, and daily context across product, engineering, QA, and marketing streams. This skill applies deterministic task management through scripted work-manager operations and handoff documents to ensure serialized, traceable workflows.

What's the best way to manage isolated worktrees for concurrent AI agent development?

Isolated worktrees for concurrent AI agent development are managed by creating branch workspaces through a work-manager script. The script initializes the worktree, sets up the context file, and adds tasks to ensure deterministic task tracking and isolation.

How do I generate daily logs and handoff documents for AI agent context propagation?

Daily logs and handoff documents for AI agent context propagation are generated automatically to track the lifecycle of operations across teams. This ensures seamless daily-log handoffs and maintains context across product, engineering, QA, and marketing streams.

Can I use scripted task management to serialize workflows across multiple AI agents?

Yes, scripted task management serializes workflows across multiple AI agents by applying deterministic work-manager operations. It uses context files and handoff documents to ensure every task is traceable and operations flow sequentially across all teams.

When do I need CEO-led orchestration for AI agent branches and daily logs?

CEO-led orchestration for AI agent branches and daily logs is needed when coordinating multiple sub-agents across distinct team streams like product, engineering, QA, and marketing. It ensures traceable, serialized workflows through worktree creation and context handoffs.

Does async agent orchestration work without external dependencies?

Yes, async agent orchestration works without external dependencies by relying on internal scripts and context files. The skill uses a work-manager script to handle branch workspaces, task additions, and handoff document generation deterministically.