Skip to main content
Renderers transform compiled document output into specific formats. While all Colin templates are written in Markdown with Jinja syntax, the final output can be JSON, YAML, or plain Markdown depending on the renderer specified. Colin includes three built-in renderers:
RendererExtensionDescription
Markdown.mdDefault passthrough renderer
JSON.jsonParses markdown structure into JSON
YAML.yamlValidates YAML output

​
Configuration

Specify the output format in document frontmatter: 
---
colin:
  output:
    format: json
---

## name
John Doe

## age
```json
30


This produces `document.json` with content `{"name": "John Doe", "age": 30}`.

You can also specify a custom output path:

```yaml
colin:
  output:
    format: json
    path: config/user.json  # writes to output/config/user.json
See Document Settings for full output configuration options.

​
How Renderers Work

Renderers operate after Jinja template rendering completes. The pipeline is: 
Source (.md) → Jinja2 rendering → Raw output string → Renderer → Final file
Each renderer can:
  • Transform content: Convert between formats (e.g., markdown headers to JSON keys)
  • Validate output: Ensure content is valid for the target format
  • Set file extension: Determine the output filename extension

​
Default Behavior

Without an output specification, Colin uses the Markdown renderer which passes content through unchanged. This is ideal for generating documentation or reports. 
---
title: My Report
---

# Summary

{{ ref('data').content | llm_extract('key findings') }}

​
Choosing a Renderer

Use Markdown (default) when:
  • Generating documentation or reports
  • The output is meant for human reading
  • You want to preserve formatting
Use JSON when:
  • Building structured data for APIs or agents
  • Creating configuration files
  • Output will be consumed by programs
Use YAML when:
  • Creating human-readable configuration
  • Output needs to preserve comments and structure