work-session

Coordinate multi-task AI development sessions with configurable human checkpoints.

8|2|Updated Jan 15, 2026
One-click install
npx skills add https://github.com/srstomp/pokayokay --skill work-session
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: work-session
Source: https://github.com/srstomp/pokayokay/tree/main/plugins/pokayokay/skills/work-session
Command: npx skills add https://github.com/srstomp/pokayokay --skill work-session

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

The Work Session skill addresses the complexity of coordinating AI-assisted development across multiple agents and sessions, ensuring coherent context, controllable human checkpoints, and continuity between tasks.

Core Features & Use Cases

  • Fresh task context via subagent dispatch to prevent context bleed between tasks.
  • Configurable checkpoint modes (supervised, semi-auto, auto, unattended) for precise human oversight.
  • Session continuity across conversations via MCP and ohno integration.
  • Smart worktree and isolation decisions based on task type to optimize productivity.

Quick Start

Initialize the MCP by running npx pokayokay and start a supervised work session.

Frequently Asked Questions about work-session

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

FAQPage Schema
How do I orchestrate AI work sessions with human checkpoints?

To coordinate AI-assisted development across multiple agents, you can use configurable checkpoint modes like supervised, semi-auto, auto, and unattended, ensuring coherent context and continuity between tasks.

What is the best way to prevent context bleed between AI development tasks?

The best way to prevent context bleed between tasks is by dispatching subagents to establish fresh task context, isolating multi-task execution while preserving session continuity.

How do I maintain session continuity across AI conversations?

You can maintain session continuity across AI conversations by integrating MCP and ohno to coordinate subagents, preserving context and ensuring safe handoffs across multiple tasks.

Can I configure unattended execution modes for AI-assisted development?

Yes, you can configure unattended execution modes for AI-assisted development, alongside supervised, semi-auto, and auto modes, allowing precise human oversight and governance for multi-task workflows.

When do I need worktree isolation for AI agent workflows?

You need worktree isolation for AI agent workflows when making smart decisions based on task type, optimizing productivity while ensuring safe progression and isolation during complex multi-task sessions.

How to start a supervised work session with MCP integration?

To start a supervised work session with MCP integration, initialize the MCP by running npx pokayokay, setting up the environment to coordinate subagents and apply configurable governance checkpoints.