Update README.md

n8n-master-corpus
The n8n-master-corpus is a massive, curated dataset of 36,405 labeled n8n automation workflows. It serves as a definitive resource for training LLMs
on workflow generation, analyzing automation patterns, or quickly finding ready-to-use templates for the n8n ecosystem.

🚀 Highlights
- Scale: 36,405 unique workflows.
- Copy-Paste Ready: The workflow_json column contains the exact JSON structure required by n8n.
- Diversity: Includes workflows ranging from simple logic to complex AI-Agent integrations and MCP server implementations.
- Labeled: Metadata includes node counts, integration types, and categories.

📊 Dataset Structure

Each entry in the dataset contains the following fields:


┌───────────────┬─────────────────────────────────────────────────────────────────────────────────────────────────────────────┐
│ Field │ Description │
├───────────────┼─────────────────────────────────────────────────────────────────────────────────────────────────────────────┤
│ workflow_id │ Original unique identifier for the workflow. │
│ workflow_name │ Descriptive name of the automation. │
│ node_count │ Number of nodes within the workflow. │
│ integrations │ JSON list of third-party services used (e.g., Slack, Google Sheets, OpenAI). │
│ category │ The thematic pack or source the workflow belongs to. │
│ workflow_json │ The core data. A stringified JSON object compatible with n8n's "Import from File" or "Paste" functionality. │
└───────────────┴─────────────────────────────────────────────────────────────────────────────────────────────────────────────┘

🛠️ How to Use

In Python (Datasets Library)
- Load the dataset using the datasets library.
- Access any row and parse the workflow_json string using the standard json library.
- Use the resulting dictionary as needed for training or analysis.

In n8n
- Obtain the workflow_json string for any entry.
- In your n8n canvas, press Ctrl + V (or Cmd + V) and paste the JSON string directly.
- The workflow will instantly manifest with all nodes and connections intact.

📁 Source & Provenance
This dataset is a collection of curated community packs, synthetic generations, and lab-tested AI automations. It includes workflows from the AI
Automation Lab, Community Mega-Pack, and specialized Synapse Pro-Ops collections.

⚖️ License
This dataset is released under the Apache-2.0 license. Individual workflow logic remains subject to the licenses of their original respective authors
where applicable.

README.md

@@ -24,4 +24,19 @@ configs:
   data_files:
   - split: train
     path: data/train-*
----
+license: apache-2.0
+task_categories:
+- text-generation
+- feature-extraction
+language:
+- en
+tags:
+- n8n
+- automation
+- workflow
+- low-code
+- agents
+- mcp
+size_categories:
+- 10K<n<100K
+---