Overview
Placeholders pass data between blocks in your workflow. They connect user inputs, step outputs, and brand information throughout the execution sequence. Three types of placeholders:- Input Placeholders: Values defined by the user in the Input block
- Step Output Placeholders: Data returned by previous workflow steps
Input Placeholders
Input placeholders are defined in your workflow’s Input Block. Syntax:Example
If your Input Block defines a placeholder calledkeyword, reference it in any later step:
topic, url, audience). They automatically appear in the placeholder selector inside each block.
Step Output Placeholders
Each step generates an output you can reference in later steps. Syntax:<n> is the step number (step_1, step_2, etc.).
Example
If Step 1 is a Google Search block and Step 2 is an LLM block, reference Step 1’s results in your Step 2 prompt:Accessing Structured Data
If a step returns structured data (JSON), access specific properties:Best Practices
- Use the Placeholder Selector ({ } icon on the top left) to insert placeholders without typing manually
- Check the Test Panel to preview step outputs before referencing them
- For JSON outputs, expand the result in the Test Panel to confirm the correct structure (keys, arrays)
- Placeholders can only reference previous steps, not future ones
Common Use Cases
| Goal | Example |
|---|---|
| Pass user input into an AI prompt | Generate a blog outline for {{input.keyword}}. |
| Include brand tone or persona | Write this in {{input.brandkit.writingSample.outLine}} style. |
| Reuse output from a previous step | Summarize {{step_1.output}} into three key points. |
| Extract specific data from an API response | Analyze {{step_3.output.results[0].url}}. |
| Combine results for formatted output | The final headline is {{step_2.output.headline}}. |
What’s Next
Now that you understand placeholder referencing:- See placeholders in action in Building Workflows
- Learn about block types in Advanced Features
- Explore real examples in Examples & Tutorials