DAGCLI Docs

dag node

Discover and inspect the node types available in your installation. Use these subcommands to explore what each node does, what ports it exposes, and how to wire it into a pipeline.

dag node list

Print all registered node types grouped by category.

bash
dag node list
text
Core
  input             Injects --input values into the pipeline
  text-output       Prints the text port to stdout
  text-template     Renders a Handlebars template with upstream port values

Provider
  llm-text-anthropic   Anthropic Claude text completion
  llm-text-openai      OpenAI Chat Completions text completion
  llm-text-google      Google Gemini text completion

dag node info <type>

Show the input ports, output ports, and config keys for a specific node type.

bash
dag node info llm-text-anthropic
text
Node: llm-text-anthropic
Category: Provider

Input ports
  text    string   The prompt text to send to the model

Output ports
  text    string   The model's text completion

Config keys
  model       string   Model ID (default: claude-3-5-haiku-20241022)
  maxTokens   number   Maximum tokens in the response (default: 1024)
  temperature number   Sampling temperature 0–1 (default: 0.7)

dag node example <type>

Generate a minimal, runnable DAG JSON file that exercises the specified node type. The output is printed to stdout and can be piped into a file or passed directly to dag run.

bash
dag node example llm-text-anthropic
json
{
  "dagId": "llm-text-anthropic-example",
  "nodes": [
    { "id": "input",  "type": "input" },
    { "id": "llm",    "type": "llm-text-anthropic" },
    { "id": "output", "type": "text-output" }
  ],
  "edges": [
    { "source": "input", "target": "llm"    },
    { "source": "llm",   "target": "output" }
  ]
}

Save and run immediately:

bash
dag node example llm-text-anthropic > example.dag.json
dag run example.dag.json --input text="Why is the sky blue?"

Available Node Types

TypeCategoryDescription
inputCoreInjects --input values into the pipeline.
text-outputCorePrints the text port to stdout.
text-templateCoreRenders a Handlebars template with upstream port values.
llm-text-anthropicProviderCalls the Anthropic Claude API and returns a text completion.
llm-text-openaiProviderCalls the OpenAI Chat Completions API and returns a text completion.
llm-text-googleProviderCalls the Google Gemini API and returns a text completion.