agentic-patterns

Design and implement multi-agent workflows with pattern strategies and smoke tests.

1|Updated Apr 3, 2026
One-click install
npx skills add https://github.com/bflandev/threadline-ai --skill agentic-patterns-bflandev
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agentic-patterns
Source: https://github.com/bflandev/threadline-ai/tree/main/skills/agentic-patterns
Command: npx skills add https://github.com/bflandev/threadline-ai --skill agentic-patterns-bflandev

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Design and implement multi-agent workflows using proven pattern strategies.

Core Features & Use Cases

  • Pattern catalog covering single-agent, sequential, parallel, loop, coordinator, hierarchical, swarm, ReAct, human-in-the-loop, and custom logic
  • Builder templates and step-by-step guidance for pattern construction
  • Wiring instructions detailing how to connect components with Claude Code's Agent tool
  • Validation criteria and smoke tests to verify correct pattern wiring

Quick Start

Begin with a simple single-agent pattern to establish the workflow, then extend to multi-agent orchestrations.

Frequently Asked Questions about agentic-patterns

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

FAQPage Schema
What are the standard multi-agent workflow patterns for orchestrating AI agents?

Standard multi-agent workflow patterns include single-agent, sequential, parallel, loop, coordinator, hierarchical, swarm, ReAct, and human-in-the-loop designs. These pattern strategies provide structured approaches for designing complex agent orchestrations and coordinating tasks across multiple agents.

How do I wire multiple agents together using a coordinator or hierarchical pattern?

You can wire multi-agent patterns using provided builder templates and step-by-step wiring instructions. These templates detail how to connect components correctly, enabling you to construct hierarchical or coordinator workflows that manage task delegation across agents effectively.

When should I use a swarm pattern versus a sequential workflow for my agent orchestration?

Sequential workflows execute agent tasks in a linear chain, making them ideal for predictable, step-by-step processing. Swarm patterns allow agents to operate dynamically and collaboratively, making them better suited for complex, unpredictable tasks requiring flexible agent coordination and parallel problem-solving.

How can I verify that my multi-agent pattern is wired correctly and functioning as expected?

You can verify correct pattern wiring by applying provided validation criteria and running practical smoke tests. These verification methods check your multi-agent workflow construction, ensuring the agents are properly connected and the orchestration logic functions as designed before deployment.

Can I implement a human-in-the-loop logic pattern within my automated agent workflow?

Yes, human-in-the-loop is included as a specific pattern in the multi-agent workflow catalog. It allows you to design orchestrations that pause for human review, input, or approval at defined steps, combining automated agent processing with manual oversight.

Does this approach support building custom logic patterns for specialized agent workflows?

Yes, custom logic is explicitly supported alongside standard patterns like ReAct and hierarchical designs. You can use the builder templates to construct specialized agent workflows tailored to your specific orchestration requirements beyond the predefined pattern catalog.