Crabflow 🦀

Crabflow is a powerful tool for running REST workflows. It allows you to define and execute sequences of HTTP requests with dependencies, retries, and response validation.

• Crabflow 🦀
  • Features ✨
  • Installation 📥
    • Homebrew (MacOS/Linux (x86_64) - x86_64/arm64)
    • Podman (Linux x86_64)
    • Download and Run Binary
    • Build and Run Binary
  • Usage 🚀
  • Workflow Configuration ⚙️
    • Task Configuration
    • Response Expectations
    • Environment Variables
  • License

Features ✨

• 📝 Define workflows in YAML format
• 🌐 Execute HTTP requests with various methods (GET, POST, etc.)
• 📦 Support for different body types (JSON, form-urlencoded, raw, multipart)
• ✅ Response validation using status codes, JSON paths, and raw text matching
• 🔧 Environment variable resolution
• 🔄 Response registration and reference between tasks
• 🔁 Automatic retries with configurable delay
• 🔐 Basic authentication support
• 🎯 Custom headers support

Installation 📥

Homebrew (MacOS/Linux (x86_64) - x86_64/arm64)

brew install stenstromen/tap/crabflow

Podman (Linux x86_64)

podman run --rm 
-e PASSWD=123 \
-e X_API_KEY=123 \
-e FOO=bar \
-v /path/to/workflow.yaml:/data/workflow.yaml:ro \
ghcr.io/stenstromen/crabflow:latest /crabflow /data/workflow.yaml

Download and Run Binary

• For MacOS and Linux: Checkout and download the latest binary from Releases page
• For Windows: Build the binary yourself.

Build and Run Binary

cargo build --release
./target/release/crabflow

Usage 🚀

Create a workflow file (e.g., workflow.yaml) and run it:

crabflow workflow.yaml

If no workflow file is specified, Crabflow will look for workflow.yaml in the current directory.

Workflow Configuration ⚙️

Here's an example workflow configuration:

name: example-pipeline
tasks:
  - name: fetch-users
    type: http
    method: GET
    url: http://api.example.com/users
    retries: 2
    expect:
      type: JsonPath
      path: data.users
      value: "[]"

  - name: create-user
    type: http
    method: POST
    url: http://api.example.com/users
    depends_on: [fetch-users]
    body_type: json
    body:
      name: "John Doe"
      email: "john@example.com"
    expect:
      - type: Status
        code: 201
      - type: JsonPath
        path: "id"
        value: "123"
    register: new_user

  - name: verify-user
    type: http
    method: GET
    url: http://api.example.com/users/{{new_user.json.id}}
    depends_on: [create-user]
    expect:
      - type: Status
        code: 200
      - type: JsonPath
        path: "name"
        value: "John Doe"

  - name: display-user
    type: display
    variables:
      - new_user

Task Configuration

Each task in the workflow can have the following properties:

• name: Unique identifier for the task
• type: Type of task (currently only "http" is supported)
• method: HTTP method (GET, POST, PUT, DELETE, etc.)
• url: Target URL
• headers: Custom HTTP headers
• body: Request body (optional)
• body_type: Type of body (json, form-urlencoded, raw, form-multipart)
• depends_on: List of task names that must complete before this task
• retries: Number of retry attempts (default: 1)
• retry_delay: Delay between retries in seconds (default: 5)
• expect: List of expectations for the response
• register: Name to register the response for reference in other tasks
• auth: Basic authentication credentials

Response Expectations

You can validate responses using:

• Status: Expected HTTP status code
• JsonPath: Expected value at a JSON path
• Raw: Expected text in the response

Environment Variables

Use environment variables in your workflow:

url: http://api.example.com/{{env.API_VERSION}}/users
headers:
  Authorization: "Bearer {{env.API_KEY}}"

License

MIT License