feat: add data utility functions

[CristianAmbrosini]

Test PR to validate that org review guidelines are injected via ~/CLAUDE.md and picked up by the review agent.

[sonar-review-dev18]

Summary

This PR adds a new data_utils.py module with four utility functions for common data manipulation tasks:

• normalize(): Min-max normalization scaling values to [0, 1]
• flatten(): Recursively flattens nested lists into a single level
• chunk(): Splits a list into fixed-size sublists
• merge_dicts(): Recursively merges two dictionaries with special handling for nested dict values

The PR is described as a test to validate that org review guidelines (from ~/CLAUDE.md) are properly injected and picked up by the review agent.

What reviewers should know

What to review:

• The four utility function implementations in data_utils.py
• Consider whether these functions are adequately tested (none currently included)
• Check if there's a plan to integrate these into the codebase or if they're standalone examples

Things to note:

• No unit tests are included with this addition
• Functions lack docstrings—may want to confirm if that aligns with project standards
• This is a new module that will need to be imported/used somewhere to be effective
• The normalize() function will raise an exception if all values are identical (division by zero)

---

• Generate Walkthrough
• Generate Diagram

🗣️ Give feedback

[sonarqube-cloud-dev18]

SonarQube reviewer guide

Quality Gate failed

Failed conditions
0.0% Coverage on New Code (required ≥ 80%)

See analysis details on SonarQube Cloud

[sonar-review-dev18]

There are two real crash bugs in normalize() that need to be fixed before merge.

SonarQube Cloud status: Coverage on new code is 0% against an 80% threshold — see the reviewer guide above for details. The coverage gap is meaningful here: the uncovered edge cases in normalize() are exactly the paths that contain bugs.

🗣️ Give feedback

[sonar-review-dev18]

Missing test coverage: normalize() has two unguarded crash paths that no test currently exercises:

• Empty list — min([]) raises ValueError. There is no guard before calling min/max.
• All-identical values — when every element is the same, mx - mn == 0, causing ZeroDivisionError on the return line.

Both will blow up in production on valid-looking inputs. The uniform-values case is especially sneaky because a caller has no obvious reason to expect a crash.

A minimal fix for the zero-division case:

Suggested change

	def normalize(values):
	mn = min(values)
	mx = max(values)
	return [(v - mn) / (mx - mn) for v in values]
	def normalize(values):
	mn = min(values)
	mx = max(values)
	if mx == mn:
	return [0.0] * len(values)
	return [(v - mn) / (mx - mn) for v in values]

• Mark as noise

[sonar-review-dev18]

Missing test coverage: chunk(lst, 0) raises ValueError: range() arg 3 must not be zero with no helpful message. Since size comes from the caller, a zero or negative value is a plausible mistake. Add a guard and a clear error:

if size <= 0:
    raise ValueError(f"size must be a positive integer, got {size}")

• Mark as noise