workflow

Orchestrate predefined AI coding workflows by parsing workflow.md and advancing through stages.

13|4|Updated Mar 17, 2026

Search Tags:#workflow #ai #agents #orchestration #stages #coordinator #memorizer

Ready to run

npx skills add https://github.com/ob-labs/talos --skill workflow-ob-labs

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: workflow
Source: https://github.com/ob-labs/talos/tree/main/skills/workflow
Command: npx skills add https://github.com/ob-labs/talos --skill workflow-ob-labs

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

通过读取 workflow.md 并推进阶段，该 Skill 实现对预定义 AI 编码工作流的协调执行，确保协调者角色的职责、代理委托和安全约束得到遵循，同时避免在执行阶段对项目源码进行写操作。

Core Features & Use Cases

基于 workflow.md 的阶段化工作流编排与执行。
在需要时将任务委托给 subagent，并强制执行协调规则与安全约束。
提供从初始化到完成的完整生命周期，以及 memorizer 记忆功能的交互接口。

Quick Start

在 Claude Code 中输入 /workflow <name> 启动相应工作流。

Frequently Asked Questions about workflow

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

FAQPage Schema

How do I orchestrate predefined AI coding workflows stage by stage?▼

AI workflow orchestration parses a workflow.md file to advance through predefined stages. It enforces coordinator rules, delegates tasks to subagents, and validates each stage without writing to project source code during execution.

How do I start a named AI agent workflow in Claude Code?▼

Start a named AI workflow by entering /workflow <name> in Claude Code. This initializes the orchestration lifecycle, applying memorizer memory prompts and stage validation to execute the predefined coding workflow.

Can I delegate workflow orchestration tasks to subagents?▼

Yes, the workflow coordinator delegates tasks to subagents while enforcing safety constraints. Subagents execute assigned stages, but the system prevents writing to project source code during the execution phase to maintain safety.

What are the limitations of using workflow.md for AI agent orchestration?▼

A key limitation is that the system avoids writing to project source code during execution phases. Orchestration relies entirely on parsing predefined workflow.md files, requiring strict adherence to memorizer memory prompts and stage validation rules.

Does AI workflow orchestration support memorizer memory prompts?▼

Yes, the workflow orchestration system supports memorizer memory prompts. It specifies requirements and constraints for these prompts to interact with the memorizer, maintaining state throughout the workflow lifecycle from initialization to completion.