# Build Streamed Multi-Agent Workflows

Sequential AI workflows often lose context between steps. This skill defines stream-chain patterns that pass each result into the next task.

## Install

```bash
npx skillstore add dnyoussef/stream-chain
```

## Metadata

- Status: approved
- Slug: dnyoussef-stream-chain
- Version: 1.0.0
- Author version: 1.0.0
- Skillstore revision: r1
- Version status: valid
- Tree hash: daf7c8eff1b3855c8c65519ec76b798127113782b08642871a1724244a424734
- Author: DNYoussef
- GitHub username: DNYoussef
- License: MIT
- Repository: https://github.com/DNYoussef/ai-chrome-extension/tree/main/.claude/skills/stream-chain
- Ref: 72d5025b022c77f7a51bdf5c1637c689c80e89d1
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands
- Quality score: 80
- Quality tier: silver
- Public page: https://skillstore.pages.dev/skills/dnyoussef-stream-chain
- Manifest: https://skillstore.pages.dev/api/skills/dnyoussef-stream-chain/manifest

## Capabilities

- Defines custom stream-chain workflows with two or more sequential prompts.
- Documents predefined pipelines for analysis, refactoring, testing, and optimization.
- Shows how prior step output becomes context for later tasks.
- Explains timeout, verbose, and debug options for chain execution.
- Provides examples for code review, migration, security review, and data transformation.

## Use Cases

- Plan a feature delivery chain: Create a sequence that analyzes requirements, designs architecture, implements code, writes tests, and documents the result.
- Audit code quality systematically: Run analysis, refactor, test, and optimization steps as a repeatable review workflow.
- Transform project data through stages: Structure extraction, normalization, validation, enrichment, and reporting tasks as a chained data workflow.

## Prompt Templates

### Start a simple chain

```
Create a stream-chain workflow that analyzes a small module, suggests improvements, and writes a short action plan.
```

### Use a predefined pipeline

```
Show how to run the analysis pipeline, explain each stage, and describe the expected final report.
```

### Design a custom delivery workflow

```
Build a five-step stream-chain plan for a new API feature, including design, implementation, tests, documentation, and review.
```

### Tune an advanced migration chain

```
Create an advanced migration chain with validation steps, timeout guidance, debug usage, and a final rollback checklist.
```

## Limitations

- Requires a working claude-flow stream-chain command in the user environment.
- Does not include executable implementation files inside the skill package.
- Outputs depend on the connected agent models and project context.
- Long or complex chains may need higher per-step timeout values.

## Best Practices

- Use specific prompts so each chain step has a measurable result.
- Place verification steps after implementation or transformation steps.
- Increase timeouts for analysis, migration, or large codebase workflows.

## Anti Patterns

- Do not use vague prompts such as check everything or improve code.
- Do not chain unrelated tasks that need separate context or review.
- Do not skip validation when a chain changes code or data structures.

## Security Audit

- Audited at: 2026-07-06T11:39:01.445\+00:00
- Summary: All 87 static findings are false positives caused by Markdown inline code and fenced examples in SKILL.md. The skill is documentation for stream-chain workflows and does not contain executable scripts, network calls, credential access, or prompt-injection language. No semantic security issues were found.

## Stats

- Views: 309
- Downloads: 8
- Favorites: 0
- Popularity score: 0