Simple Sequential Workflow¶

A basic workflow demonstrating sequential step execution with data passing.

Overview¶

This example shows:

Basic workflow structure
Sequential step dependencies
Passing data between steps with {{steps.X.output}}
Simple inline agents

The Workflow¶

examples/simple-sequential.yaml

name: "simple-sequential"
description: "Sequential pipeline with three review stages"

inputs:
  files:
    description: "Files to review (glob pattern)"
    default: "src/**/*.go"

agents:
  security-reviewer:
    inline:
      description: "Reviews code for security issues"
      prompt: |
        You are a security expert. Analyze code for vulnerabilities.
        Focus on injection, authentication, and data exposure issues.
        Be concise and cite specific locations.

  performance-reviewer:
    inline:
      description: "Reviews code for performance issues"
      prompt: |
        You are a performance expert. Analyze code for efficiency.
        Focus on algorithmic complexity, memory usage, and I/O.
        Be concise and cite specific locations.

  aggregator:
    inline:
      description: "Combines multiple reviews"
      prompt: |
        You summarize multiple code reviews into a single report.
        Prioritize findings by severity and actionability.

steps:
  # Step 1: Security review
  - id: security-review
    agent: security-reviewer
    prompt: "Review the code in {{inputs.files}} for security issues."

  # Step 2: Performance review (waits for security review)
  - id: perf-review
    agent: performance-reviewer
    prompt: |
      Review the code in {{inputs.files}} for performance issues.

      For context, the security review found:
      {{steps.security-review.output}}
    depends_on: [security-review]

  # Step 3: Summary (waits for performance review)
  - id: summary
    agent: aggregator
    prompt: |
      Create an executive summary combining:

      ## Security Findings
      {{steps.security-review.output}}

      ## Performance Findings
      {{steps.perf-review.output}}

      Provide: top priorities, quick wins, and recommendations.
    depends_on: [perf-review]

output:
  steps: [summary]
  format: markdown

How It Works¶

security-review → perf-review → summary

security-review runs first (no dependencies)
perf-review waits for security-review, then runs with its output as context
summary waits for perf-review, then combines both outputs

Running the Example¶

Mock Mode (No API Calls)¶

goflow run \
  --workflow examples/simple-sequential.yaml \
  --inputs files='pkg/**/*.go' \
  --mock \
  --verbose

Expected output:

[INFO] Loading workflow: examples/simple-sequential.yaml
[INFO] Starting workflow: simple-sequential
[INFO] Step 1/3: security-review (security-reviewer)
[INFO] Step 2/3: perf-review (performance-reviewer)
[INFO] Step 3/3: summary (aggregator)
[INFO] Workflow completed

## Step: summary

mock output

Real Mode¶

goflow run \
  --workflow examples/simple-sequential.yaml \
  --inputs files='pkg/workflow/*.go' \
  --verbose

Key Concepts Demonstrated¶

1. Input Parameters¶

inputs:
  files:
    description: "Files to review"
    default: "src/**/*.go"

Override at runtime: --inputs files='other/*.go'

2. Step Dependencies¶

- id: perf-review
  depends_on: [security-review]  # Must wait

3. Template Variables¶

prompt: |
  Context from security:
  {{steps.security-review.output}}

Injects the actual output from the referenced step.

Variations¶

Adding More Stages¶

steps:
  - id: security-review
    ...

  - id: perf-review
    depends_on: [security-review]

  - id: style-review  # New step
    agent: style-reviewer
    depends_on: [perf-review]

  - id: summary
    depends_on: [style-review]  # Now waits for style

Using File-Based Agents¶

Replace inline agents with files:

agents:
  security-reviewer:
    file: "./agents/security-reviewer.agent.md"