perf(data_loader): cache JSON data in memory to avoid repeated disk reads

[Alm0stSurely]

Problem

As described in #11, utils/data_loader.py currently reads projects.json from disk on every call to load_all_projects().

Solution

Implemented in-memory cache with module-level variable:

• _projects_cache stores data after first read
• clear_cache() for test isolation
• Updated tests to reset cache in setup_module()

Benchmarks

• Cached: 0.06 µs per call
• Uncached: 51.84 µs per call
• Speedup: 847x

Testing

All 25 tests pass.

Fixes #11

[komalharshita]

This is a solid improvement overall.

The caching logic is simple, clean, and fits the project really well. I also like that you added clear_cache() for test isolation and included benchmark results to show the performance improvement. The code stays beginner-friendly while still improving efficiency, which is exactly the direction we want for DevPath.

I reviewed the updated files and everything looks good to merge.

[komalharshita]

Nice work on this optimization.
Reading the JSON file on every request was unnecessary, and the in-memory cache solves that cleanly without adding extra complexity. The benchmark script and test updates were a great addition too.
Thanks for keeping the implementation readable and well-documented.