takt

Orchestrate multi-agent teams across piece YAML workflows with auditable prompts.

1.3k|86|Updated Jan 25, 2026

Search Tags:#workflow #multi-agent #orchestration #yaml #prompts #policies #takt

Ready to run

npx skills add https://github.com/nrslib/takt --skill takt

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: takt
Source: https://github.com/nrslib/takt/tree/main/builtins/skill
Command: npx skills add https://github.com/nrslib/takt --skill takt

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

TAKT orchestrates multi-agent collaboration using piece YAML workflows, providing deterministic coordination, visibility, and auditable logs for AI agent teams.

It covers human-in-the-loop interventions, policy and knowledge injection, and modular sections for prompts, personas, and instructions.

Core Features & Use Cases

End-to-end multi-agent orchestration with explicit movements, personas, and policies.
Human-in-the-loop control and full auditability for compliance and debugging.
Knowledge and policy injection to enforce domain standards across workflows.

Quick Start

Invoke TAKT with a piece YAML and run the initial movement to spawn team mates, execute steps, and progress through the workflow.

Frequently Asked Questions about takt

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

FAQPage Schema

How do I coordinate multi-agent workflows using YAML?▼

You can coordinate multi-agent workflows using YAML by defining explicit movements, personas, and policies. This approach enforces deterministic coordination and produces repeatable results across complex task scenarios.

What is the best way to add human-in-the-loop control to AI agent orchestration?▼

Adding human-in-the-loop control to AI agent orchestration requires defining explicit intervention points in a YAML workflow. This enables full auditability for compliance and debugging while maintaining deterministic agent coordination.

Can I enforce domain policies and inject knowledge into automated agent prompts?▼

Yes, you can enforce domain policies and inject knowledge into automated agent prompts. By defining modular sections for prompts, personas, and instructions, the workflow enforces domain standards consistently across all executed steps.

How do I start running a multi-agent YAML workflow?▼

To start running a multi-agent YAML workflow, invoke the orchestrator with your YAML file and run the initial movement. This spawns team mates, executes defined steps, and progresses through the orchestrated workflow.

Why do I need auditable logs for multi-agent collaboration?▼

Auditable logs for multi-agent collaboration are needed to provide visibility and ensure compliance during complex task scenarios. They allow you to debug deterministic workflows and verify that injected policies and prompts were followed correctly.