stream-chain

Stream-Chain is a specialized skill designed for complex orchestration within the Ruflo development environment. It empowers developers and AI agents to create sophisticated, multi-step workflows where the output of one agent serves as the input for the next. By utilizing a sequential flow mechanism, Stream-Chain enables seamless data transformation, codebase analysis, and automated refactoring tasks. The skill supports both custom prompt sequences and battle-tested predefined pipelines, making it an essential tool for technical debt reduction, security auditing, and performance optimization.

• Custom chaining: Allows users to run arbitrary sequences of prompts, with each step maintaining context from the previous output, facilitating iterative code generation and validation.

• Predefined pipelines: Includes optimized workflows for analysis, refactoring, testing, and performance optimization, ensuring consistent results for common engineering tasks.

• Sequential context flow: Each step in the pipeline automatically inherits context from the prior execution, allowing agents to build on previous work, such as moving from architectural analysis to test implementation.

• High-level orchestration: Designed to work within the broader Ruflo framework, allowing for multi-agent coordination, fault-tolerant execution, and integration with existing MCP servers.

• Flexible configuration: Supports command-line execution with options for verbosity, timeouts, and debug modes to assist in monitoring deep-chain performance.

• Use cases range from codebase onboarding and security vulnerability detection to complex data processing and documentation drift analysis.

• The execution model requires at least two prompts for custom runs, ensuring that the sequential processing capability is fully utilized.

• Inputs typically consist of natural language instructions or task descriptions, with outputs providing refined, multi-stage deliverable artifacts.

• Developers should be aware of step timeouts and execution time, as complex chains involving multiple agents may require higher timeout settings in environments with restricted resources.

• Users can define custom, reusable pipelines within the .claude-flow/config.json file, allowing teams to standardize their unique development workflows and share them across the organization.