Objects

Objects are the things that pipelines operate on — the inputs that flow through the graph.

Object Types

Folder

A directory tree. The pipeline reads all files under the specified path and passes their content downstream.

{
  "type": "folder",
  "data": { "path": "./src" }
}

Use cases:

• Scan an entire project directory
• Process all files matching a pattern
• Feed a codebase to an AI agent for review

Code File

A single source file. Use when you want to target a specific file for processing.

{
  "type": "code-file",
  "data": { "path": "./src/index.ts" }
}

Use cases:

• Analyze a single configuration file
• Process a specific module
• Target a known problem file

GitHub Project

A GitHub repository. The pipeline clones or references the repository and passes its content downstream.

{
  "type": "github-project",
  "data": { "owner": "forge-town", "repo": "ordine" }
}

Use cases:

• Review external repositories
• Cross-project analysis
• Dependency auditing

How Objects Flow

Objects are source nodes in the pipeline DAG. They have no incoming edges — only outgoing ones. Their content is packaged as NodeCtx and flows to downstream operations:

[folder: ./src] ──► [operation: review] ──► [output: ./results]

Multiple objects can feed into the same operation, and their content is merged before processing.

Custom Objects (via Plugins)

Plugins can introduce new object types. For example, a documentation plugin might add a wiki-page object type, or a CI plugin might add a pull-request object type.