Format code examples in markdown documentation for TalkPipe. Use when writing or editing doc examples. Covers unindented fences, standalone vs fragment intent, and skip-extract. Complements run-documentation-examples which runs the examples.
91
91%
Does it follow best practices?
Impact
93%
1.16xAverage score across 3 eval scenarios
Passed
No findings from the security scan
Format code blocks in markdown so they render correctly and can be extracted for testing. See docs/architecture/documentation-formatting.md for full guidelines.
Place the opening fence at the start of the line (no leading spaces):
```python
from talkpipe.pipe import io
pipeline = io.echo(data="hello") | io.Print()
result = pipeline.as_function(single_out=False)()**Why**: Unindented blocks are easier for extractors to parse. Indented fences (e.g. inside lists) are supported but unindented is preferred.
### Language tag
Always include the language: ` ```python ` for Python, ` ```chatterlang ` or ` ``` ` for ChatterLang. Enables syntax highlighting and lets extractors identify example types.
## Example Intent
Make it clear whether an example is:
- **Standalone runnable**: Complete code users can copy and run. Include all imports, sources, and output.
- **Explanatory fragment**: Partial code illustrating a concept. Add context like "Part of the pipeline above..." or a comment.
## Skip Extraction
Add `# skip-extract` as the first line of a block to exclude it from the doc examples runner:
```python
# skip-extract
from setuptools import setup
setup(name="myproject", ...)Use skip-extract when:
Good (standalone, unindented):
```python
from talkpipe.pipe import io
pipeline = io.echo(data="hello") | io.Print()
pipeline.as_function(single_out=False)()
```Bad (no language tag):
```
code here
```Bad (nested backticks inside block): Avoid triple backticks within the code content; they can break extraction.
After editing docs, run the doc examples to verify they execute:
pytest tests/test_doc_examples.py -vOr use the run-documentation-examples skill. Skip LLM examples when Ollama unavailable: pytest tests/test_doc_examples.py -m "not requires_ollama" -v
sandialabs/talkpipe
3182ce0
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.