AWF — AI Workflow Framework CLI
AWF — AI Workflow Framework CLI

Workflow Examples

Real-world workflow examples demonstrating AWF capabilities.

Hello World

The simplest possible workflow:

name: hello
version: "1.0.0"
description: A simple hello world workflow

states:
  initial: greet
  greet:
    type: step
    command: echo "Hello, World!"
    on_success: done
  done:
    type: terminal
awf run hello

With Inputs

Accept user input and use it in commands:

name: greet-user
version: "1.0.0"

inputs:
  - name: name
    type: string
    required: true
  - name: greeting
    type: string
    default: "Hello"

states:
  initial: greet
  greet:
    type: step
    command: echo "{{.inputs.greeting}}, {{.inputs.name}}!"
    on_success: done
  done:
    type: terminal
awf run greet-user --input name=Alice --input greeting=Hi

Code Analysis with Agent Step

Analyze code using AI with the agent step type:

name: analyze-code
version: "1.0.0"

inputs:
  - name: file
    type: string
    required: true
    validation:
      file_exists: true

states:
  initial: read_file

  read_file:
    type: step
    command: cat "{{.inputs.file}}"
    on_success: analyze
    on_failure: error

  analyze:
    type: agent
    provider: claude
    prompt: |
      Review this code and suggest improvements:

      {{.states.read_file.Output}}
    output_format: json
    options:
      model: claude-sonnet-4-20250514
    timeout: 120
    on_success: done
    on_failure: error

  done:
    type: terminal

  error:
    type: terminal
    status: failure
awf run analyze-code --input file=main.go

Parallel Execution

Run multiple steps concurrently:

name: parallel-build
version: "1.0.0"

states:
  initial: build_all

  build_all:
    type: parallel
    parallel:
      - lint
      - test
      - build
    strategy: all_succeed
    max_concurrent: 3
    on_success: deploy
    on_failure: error

  lint:
    type: step
    command: golangci-lint run

  test:
    type: step
    command: go test ./...

  build:
    type: step
    command: go build ./cmd/...

  deploy:
    type: step
    command: echo "All checks passed, deploying..."
    on_success: done
    on_failure: error

  done:
    type: terminal

  error:
    type: terminal
    status: failure

Loop Over Files

Process multiple files using for_each:

name: process-files
version: "1.0.0"

states:
  initial: process_loop

  process_loop:
    type: for_each
    items: '["file1.txt", "file2.txt", "file3.txt"]'
    max_iterations: 100
    body:
      - process_single
    on_complete: aggregate

  process_single:
    type: step
    command: |
      echo "Processing {{.loop.Item}} ({{.loop.Index1}}/{{.loop.Length}})"
      wc -l "{{.loop.Item}}"
    on_success: process_loop
    on_failure: error

  aggregate:
    type: step
    command: echo "Processed all files"
    on_success: done
    on_failure: error

  done:
    type: terminal

  error:
    type: terminal
    status: failure

Conditional Branching

Branch based on conditions:

name: conditional-deploy
version: "1.0.0"

inputs:
  - name: env
    type: string
    required: true
    validation:
      enum: [dev, staging, prod]

states:
  initial: check_env

  check_env:
    type: step
    command: echo "Deploying to {{.inputs.env}}"
    transitions:
      - when: "inputs.env == 'prod'"
        goto: prod_approval
      - when: "inputs.env == 'staging'"
        goto: staging_deploy
      - goto: dev_deploy

  prod_approval:
    type: step
    command: echo "Requesting production approval..."
    on_success: prod_deploy
    on_failure: error

  prod_deploy:
    type: step
    command: ./deploy.sh prod
    timeout: 300
    on_success: done
    on_failure: error

  staging_deploy:
    type: step
    command: ./deploy.sh staging
    timeout: 180
    on_success: done
    on_failure: error

  dev_deploy:
    type: step
    command: ./deploy.sh dev
    timeout: 60
    on_success: done
    on_failure: error

  done:
    type: terminal

  error:
    type: terminal
    status: failure
awf run conditional-deploy --input env=prod

Retry with Backoff

Retry flaky operations:

name: api-call
version: "1.0.0"

states:
  initial: fetch_data

  fetch_data:
    type: step
    command: curl -f https://api.example.com/data
    retry:
      max_attempts: 5
      initial_delay: 1s
      max_delay: 30s
      backoff: exponential
      multiplier: 2
      jitter: 0.1
      retryable_exit_codes: [1, 22]
    on_success: process
    on_failure: error

  process:
    type: step
    command: echo "Processing data..."
    on_success: done
    on_failure: error

  done:
    type: terminal

  error:
    type: terminal
    status: failure

HTTP API Integration

Fetch data from a REST API and process the response:

name: api-integration
version: "1.0.0"

inputs:
  - name: api_token
    type: string
    required: true

states:
  initial: fetch_users

  fetch_users:
    type: operation
    operation: http.request
    inputs:
      method: GET
      url: "https://api.example.com/users"
      headers:
        Authorization: "Bearer {{.inputs.api_token}}"
        Accept: "application/json"
      timeout: 10
    on_success: create_report
    on_failure: error

  create_report:
    type: operation
    operation: http.request
    inputs:
      method: POST
      url: "https://api.example.com/reports"
      headers:
        Content-Type: "application/json"
        Authorization: "Bearer {{.inputs.api_token}}"
      body: '{"source": "users", "data": {{.states.fetch_users.Response.body}}}'
      timeout: 15
      retryable_status_codes: [429, 502, 503]
    retry:
      max_attempts: 3
      backoff: exponential
      initial_delay: 1s
    on_success: done
    on_failure: error

  done:
    type: terminal

  error:
    type: terminal
    status: failure
awf run api-integration --input api_token=$API_TOKEN

Nested Loops

Loops within loops with parent context:

name: nested-loops
version: "1.0.0"

states:
  initial: outer_loop

  outer_loop:
    type: for_each
    items: '["A", "B", "C"]'
    body:
      - inner_loop
    on_complete: done

  inner_loop:
    type: for_each
    items: '["1", "2", "3"]'
    body:
      - process
    on_complete: outer_loop

  process:
    type: step
    command: |
      echo "outer={{.loop.Parent.Item}} inner={{.loop.Item}}"
    on_success: inner_loop

  done:
    type: terminal

Output:

outer=A inner=1
outer=A inner=2
outer=A inner=3
outer=B inner=1
...

AI Agent Integration

Invoke AI agents (Claude, Codex, Gemini) directly in workflows:

name: code-review-with-agent
version: "1.0.0"

inputs:
  - name: file
    type: string
    required: true
    validation:
      file_exists: true

states:
  initial: read_file

  read_file:
    type: step
    command: cat "{{.inputs.file}}"
    on_success: analyze
    on_failure: error

  analyze:
    type: agent
    provider: claude
    prompt: |
      Review this code for issues and improvements:
      {{.states.read_file.Output}}
    options:
      model: claude-sonnet-4-20250514
    timeout: 120
    on_success: done
    on_failure: error

  done:
    type: terminal

  error:
    type: terminal
    status: failure
awf run code-review-with-agent --input file=main.go

Multi-Turn Agent Conversation

Chain multiple agent steps for conversational workflows:

name: code-conversation
version: "1.0.0"

inputs:
  - name: code
    type: string
    required: true

states:
  initial: initial_review

  initial_review:
    type: agent
    provider: claude
    prompt: |
      Review this code:
      {{.inputs.code}}
    on_success: ask_performance

  ask_performance:
    type: agent
    provider: claude
    prompt: |
      Based on your previous review:
      {{.states.initial_review.Output}}

      Now focus specifically on performance bottlenecks.
    on_success: suggest_fixes

  suggest_fixes:
    type: agent
    provider: claude
    prompt: |
      Using your analysis, suggest specific code improvements for:
      {{.inputs.code}}
    on_success: done

  done:
    type: terminal
awf run code-conversation --input code="$(cat main.py)"

Parallel AI Analysis

Run multiple agents concurrently for comprehensive analysis:

name: parallel-code-analysis
version: "1.0.0"

inputs:
  - name: code
    type: string
    required: true

states:
  initial: parallel_analysis

  parallel_analysis:
    type: parallel
    parallel:
      - security_review
      - performance_review
      - style_review
    strategy: all_succeed
    on_success: aggregate

  security_review:
    type: agent
    provider: claude
    prompt: |
      Security review:
      {{.inputs.code}}

  performance_review:
    type: agent
    provider: codex
    prompt: |
      Performance analysis:
      {{.inputs.code}}

  style_review:
    type: agent
    provider: gemini
    prompt: |
      Code style review:
      {{.inputs.code}}

  aggregate:
    type: step
    command: |
      echo "Security: {{.states.security_review.Output}}"
      echo "Performance: {{.states.performance_review.Output}}"
      echo "Style: {{.states.style_review.Output}}"
    on_success: done
    on_failure: error

  done:
    type: terminal

  error:
    type: terminal
    status: failure

Built-in Workflows

AWF includes production-ready workflows in .awf/workflows/:

WorkflowDescriptionLines
audit.yamlCode quality audit with AI analysis541
commit.yamlGit commit workflow with message generation353
feature.yamlFeature creation workflow263
implement.yamlTDD implementation workflow917

Explore these workflows for advanced patterns and best practices.

See Also

Prev
Variable Interpolation
Next
Workflow Syntax Reference