feat: python code interpreter improvements

[AngeloDanducci]

Pull Request

Issue

Fixes #1022

Description

Updates the code interpreter to allow running in sandbox with packages and allow file output - ie matplotlib

Testing

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

Attribution

• AI coding assistants used

---

Adding a new component, requirement, sampling strategy, or tool?

If your PR adds or modifies one of the types below, check the matching box. A checklist of type-specific review items will be posted as a comment.

• Component
• Requirement
• Sampling Strategy
• Tool

NOTE: Please ensure you have an issue that has been acknowledged by a core contributor and routed you to open a pull request against this repository. Otherwise, please open an issue before continuing with this pull request.

[github-actions]

This comment is managed by a bot. Editing it is fine — checking off boxes, adding notes — but please leave the HTML comment marker on the first line alone, otherwise checklist updates will break.

Tool PR Checklist

Use this checklist when adding or modifying tools in mellea/stdlib/tools/.

Protocol Compliance

• Ensure compatibility with existing backends and providers
  • For most tools being added as functions, this means that calling convert_function_to_tool works

Integration

• Tool exported in mellea/stdlib/tools/__init__.py or, if you are adding a library of tools, from your sub-module

[ajbozarth]

Some feedback from Claude — design lines up with #1022 and the helpers are deprecated cleanly. Follow-ups inline.

[AngeloDanducci]

Addressed the review feedback, also rebased on main go avoid some of the CI issues we ran into yesterday.

[ajbozarth]

Some feedback from Claude — re-reviewed, the 10 prior items are addressed. Three new RST holdouts inline (Example::, two .. deprecated::) — non-blocking, fits naturally into the docstring-sweep work item you just picked up. Worth folding "no RST in Google-style docstrings" into AGENTS.md as part of that.

[ajbozarth]

Per CONTRIBUTING.md, docstrings are Google-style — Example:: is RST. Should be a single colon (the indented block under it is already fine).

Suggested change

	Example::
	Example:

Heads up: since you just picked up the double-backtick docstring sweep, worth widening the scope to catch RST stragglers like this and the two .. deprecated:: blocks below in the same pass — and adding a line to AGENTS.md so the policy is "Google-style only, no RST" rather than just "no double backticks." Happy to defer all three to that work item.

[ajbozarth]

Same RST-vs-Google issue — .. deprecated:: is a Sphinx directive and the trailing :: on instead:: is an RST literal block. Google style uses a Deprecated: section.

Suggested change

	.. deprecated::
	Use :func:`python_tool` instead::
	Deprecated:
	Use `python_tool` instead:

Same pattern at line 1227-1228 below. Roll into the docstring sweep.

[ajbozarth]

Suggested change

	.. deprecated::
	Use :func:`python_tool` instead::
	Deprecated:
	Use `python_tool` instead:

[ajbozarth]

That review above was meant to be an approve as the docstring nits are non-blocking

[AngeloDanducci]

Sounds good, will merge as is and then pick up related doc string / AGENTS.md update in #1172