feat: add configurable chunk size

[stephantul]

This PR adds a configurable chunk size to semble. The chunk size is set to 750, which is the default today, but can be overridden by setting SEMBLE_CHUNK_SIZE. I have not documented this option yet, we don't have a great place to do so.

The PR also exposes the chunk size parameter as part of the Python API.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

Files with missing lines	Coverage Δ
src/semble/cache.py	100.00% <100.00%> (ø)
src/semble/chunking/chunking.py	100.00% <100.00%> (ø)
src/semble/index/create.py	100.00% <100.00%> (ø)
src/semble/index/index.py	100.00% <100.00%> (ø)
src/semble/types.py	100.00% <100.00%> (+2.32%)	⬆️

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[greptile-apps]

Confidence Score: 3/5

The indentation change in from_git causes token-savings stats to silently go empty for every git-indexed repository; merging as-is would ship a quiet regression on that code path.

The refactoring in from_git moved the return SembleIndex(...) outside the with tempfile.TemporaryDirectory() context manager. SembleIndex.__init__ calls _compute_file_sizes(root) synchronously, but the temp dir is already deleted at that point, so every file read silently fails and _file_sizes is always empty for git repos. This affects stats reporting on every from_git call. The rest of the changes — threading desired_chunk_length through the call stack, renaming the metadata key, env-var parsing — are clean and well-tested.

src/semble/index/index.py — the from_git method needs the return SembleIndex(...) block moved back inside the with tempfile.TemporaryDirectory() context.

_{Reviews (1): Last reviewed commit: "feat: add configurable chunk size" | Re-trigger Greptile}