stream-chain

Stream outputs from one step into the next in multi-agent workflows.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/FreakyLetsFail/open-finance --skill stream-chain-freakyletsfail
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: stream-chain
Source: https://github.com/FreakyLetsFail/open-finance/tree/main/.claude/skills/stream-chain
Command: npx skills add https://github.com/FreakyLetsFail/open-finance --skill stream-chain-freakyletsfail

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires claude-flow.

What problem does it solve?

This Skill simplifies the orchestration of multi-step AI workflows where each step's output flows into the next. It eliminates the tedious and error-prone manual chaining of individual AI tasks, enabling complex data transformations and sequential processing pipelines.

Core Features & Use Cases

  • Custom Chains: Execute custom prompt sequences with full control over each step.
  • Predefined Pipelines: Utilize battle-tested workflows for common tasks like analysis, refactoring, testing, and optimization.
  • Context Preservation: Each step receives the complete output from the previous step as context.
  • Use Case: Imagine you need to refactor a legacy codebase. Use this Skill to execute a 'refactor' pipeline that first identifies code smells, then creates a refactoring plan, and finally applies the refactoring to high-priority items, ensuring a systematic and automated approach.

Quick Start

Run a custom stream chain to 'Write a user authentication function', then 'Add input validation', and finally 'Create unit tests'.

Frequently Asked Questions about stream-chain

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

FAQPage Schema
How do I orchestrate multi-step AI workflows where each step builds on the previous output?

Stream-chain solves multi-agent workflow orchestration by streaming outputs from one step into the next. It operates in custom chains (run) and predefined pipelines (pipeline) modes, with each step receiving complete output from the previous step, supporting minimum two prompts and configurable options like verbose, timeout, and debug.

Can I use stream-chain to automate a refactoring workflow with multiple sequential steps?

Yes. Stream-chain predefined pipelines include refactoring workflows that systematically identify code smells, create refactoring plans, and apply changes to high-priority items. Context preservation ensures each step receives full output from prior steps for informed decisions.

What's the difference between custom chains and predefined pipelines in stream-chain?

Custom chains (run mode) let you execute prompt sequences with full control over each step. Predefined pipelines (pipeline mode) provide battle-tested workflows for common tasks like analysis, refactoring, testing, and optimization without manual configuration.

How does stream-chain handle context between workflow steps?

Stream-chain enforces streaming context propagation, passing complete output from each step as input to the next. This enables complex data transformations and eliminates manual chaining of individual tasks while maintaining full execution traces and stepwise results.

Do I need claude-flow to use stream-chain?

Yes. Stream-chain depends on claude-flow as a prerequisite dependency. It builds on claude-flow to enable prompt-driven orchestration and multi-step task coordination with streaming context propagation.

What's the minimum number of steps I can define in a stream-chain workflow?

Stream-chain supports a minimum of two prompts per workflow. This ensures meaningful chaining where at least one step receives and builds on output from the previous step.