docs: add decompose to tutorial with example

[AngeloDanducci]

Misc PR

Type of PR

• Bug Fix
• New Feature
• Documentation
• Other

Description

• Link to Issue: Add m decompose to tutorial #18

Testing

• Tests added to the respective file if code was changed
• New code has 100% coverage if code as added
• Ensure existing tests and github automation passes (a maintainer will kick off the github automation when the rest of the PR is populated)

[github-actions]

The PR description has been updated. Please fill out the template for your PR to be reviewed.

[mergify]

Merge Protections

Your pull request matches the following merge protections and will not be merged until they are valid.

🟢 Enforce conventional commit

Wonderful, this rule succeeded.

Make sure that we follow https://www.conventionalcommits.org/en/v1.0.0/

• title ~= ^(fix|feat|docs|style|refactor|perf|test|build|ci|chore|revert|release)(?:\(.+\))?:

[psschwei]

It might be helpful to have example commands so that folks can follow along (similar to the IVR section)

[psschwei]

Is there a smaller / simpler example we could include too? While this really shows the power of what can be done, for someone new it's a lot to try to wrap your head around.

[AngeloDanducci]

I can include a second example that is a bit more digestible for users.

[psschwei]

LGTM

cc @nrfulton in case you had any feedback

[planetf1]

Suggestion to keep our tests working nicely...

To align with the recently merged PR #372 (optimizing example testing in pytest) and ensure consistency with other examples:

1. Add Classification Marker: Add # pytest: ollama to the top of the file. This allows pytest to appropriately skip the example in environments (like CI) where Ollama is unavailable. (Consistent with simple_example.py and others).

2. Prevent Import-Time Execution: Wrap the start_session() logic in if __name__ == "__main__":. Without this, pytest executes the code immediately upon import (during the collection phase), which bypasses markers and slows down test discovery. (Consistent with simple_example.py and generative_gsm8k.py).

[AngeloDanducci]

To align with the recently merged PR #372 (optimizing example testing in pytest) and ensure consistency with other examples

These are generated outputs directly from decompose based on the prompt - mainly for users to see how you go from input prompt to intermediary to final answer.

I've updated it to match the new test structure as this will live in examples and we do have that as part of our test suite.

[jakelorocco]

This looks good! A few thoughts:

1. Can we add a flag to these examples so that they don't run? I think the goal of these examples should be to run the decompose pipeline not the decomposed mellea programs.
2. Can we add a bash script and python command in the directory so that users can clearly see that this is how to run the decompose pipeline? I think adding it to the tutorial is good but users might miss it. Ideally, if we add a python program to decompose an even smaller prompt, we can run it as part of the examples and show users how to invoke the pipeline directly from code (ie from the python functions).
3. Could you also please add a comment or docstring at the top of .py reiterating that these are the decomposed prompts not "how to decompose"?