Refly Tutorial: Build Deterministic Agent Skills and Ship Them Across APIs and Claude Code

June 15, 2026 · View on GitHub

Learn how to use refly-ai/refly to turn vibe workflows into reusable, versioned agent skills that can run via API, webhook, and CLI integrations.

Why This Track Matters

refly represents a fast-moving direction in agentic tooling: converting ad-hoc workflow logic into governed, reusable capability assets.

This track focuses on:

deploying and running Refly with a reliable local baseline
understanding the architecture behind skills, workflow runtime, and delivery paths
integrating through API, webhook, and CLI for real production usage
operating and contributing with stronger governance and observability

Current Snapshot (auto-updated)

repository: refly-ai/refly
stars: about 7.4k
GitHub release reference: v1.1.0 (checked 2026-06-15; release metadata on GitHub)

Mental Model

flowchart LR
    A[Business intent] --> B[Workflow and skill construction]
    B --> C[Deterministic runtime execution]
    C --> D[Delivery paths: API, webhook, CLI]
    D --> E[Claude Code and app integrations]
    E --> F[Governance and continuous improvement]

Chapter Guide

Chapter	Key Question	Outcome
01 - Getting Started	How do I get Refly running quickly?	Working local baseline
02 - Architecture and Component Topology	How are API, web, and packages organized?	System-level understanding
03 - Workflow Construction and Deterministic Runtime	How do I build reliable executable workflows?	Better workflow design
04 - API and Webhook Integrations	How do I integrate Refly into external systems?	Production-ready integration path
05 - Refly CLI and Claude Code Skill Export	How does CLI-driven workflow orchestration work?	Repeatable automation interface
06 - Observability, Deployment, and Operations	How do I operate Refly beyond a demo?	Operational runbook baseline
07 - Troubleshooting, Safety, and Cost Controls	How do I debug failures and reduce risk?	Reliability and guardrail playbook
08 - Contribution Workflow and Governance	How do teams contribute effectively?	Contributor readiness

What You Will Learn

how to build and run deterministic workflows in Refly
how to connect Refly to APIs, webhooks, and agent runtimes
how to use CLI workflows for structured automation and debugging
how to operate and evolve Refly in team or production contexts

Source References

Start with Chapter 1: Getting Started.

Full Chapter Map

Generated by AI Codebase Knowledge Builder