[FROWNINGdev]

Summary

Python-only release (no extension changes).

• nplusone reports loops over sliced querysets — fix(nplusone): detect sliced querysets #167 by @syedmustafan. This PR also unwraps a slice of a slice (qs[:50][:10]), as CodeRabbit suggested on fix(nplusone): detect sliced querysets #167, with a regression test that fails on the old code.
• Nested loops inherit the outer queryset's eager loading, and chain accesses suggest the full path such as select_related("product__category") — fix(nplusone): inherit outer eager loading in nested loops and suggest full relation paths #168 by @Nitjsefnie.
• Agent skill skills/django-orm-lens 1.2.0: Playbook D says the scan links inner loops to an outer prefetch from py-1.13.6, and keeps the manual checks for 1.13.5 and earlier.
• tests/test_mcp_server.py imports unittest.mock directly, closing the CodeQL py/import-and-import-from note.
• Version bumped to 1.13.6 in cli/pyproject.toml, cli/server.json, smithery.yaml and the pre-commit rev: in the READMEs.

Test plan

• cd cli && pytest -q — 1004 passed, 124 subtests
• python -m mypy django_orm_lens and python -m ruff check django_orm_lens tests — clean
• Examples from nplusone misses loops over a sliced queryset and nested relation chains #145 and nplusone still flags the inner loop after the outer queryset prefetches it #146 run on a demo project: sliced loop flagged with select_related("customer"), chain suggests product__category, fully prefetched inner loop is no longer flagged.

Closes nothing new: #145 and #146 are already closed by #167 and #168.

Summary by CodeRabbit

• Bug Fixes
  • N+1 detection now identifies loops over sliced QuerySets, including nested slices, and provides an eager-loading suggestion.
• Documentation
  • Updated pre-commit examples and guidance to reference version 1.13.6, including notes on sliced QuerySets and deeper relation checks.
• Chores
  • Updated package and integration version information to 1.13.6.

[coderabbitai]

Caution

Review failed

The pull request is closed.

Walkthrough

The scanner now unwraps nested QuerySet slices before resolving loop sources, and a regression test checks the resulting finding and eager-loading suggestion. Release metadata, pre-commit examples, changelog, and skill guidance were updated for version 1.13.6.

Changes: QuerySet Slice Detection

Layer / File(s)	Summary
Nested slice scanning and regression coverage cli/django_orm_lens/query_analyzer.py, cli/tests/test_nplusone_detector.py, cli/tests/test_mcp_server.py, CHANGELOG.md	The scanner repeatedly unwraps nested subscript nodes. A regression test checks that a twice-sliced QuerySet loop reports the author finding and suggests select_related("author"). The MCP server test now uses unittest.mock.patch.object.
Version metadata and release guidance cli/pyproject.toml, cli/server.json, smithery.yaml, README.md, cli/README.md, docs/i18n/README.*.md, skills/django-orm-lens/SKILL.md	Version metadata and pre-commit examples now use 1.13.6. The skill guidance describes scan behavior by version and adds manual checks for dynamically built or multi-function QuerySets.

Priority: Not assessed

Estimated code review effort: 2 (Simple) | ~10 minutes

Change: Bug fix

Possibly related PRs

• FROWNINGdev/django-orm-lens#167: Added the initial slice-unwrapping behavior that this change extends to nested slices.
• FROWNINGdev/django-orm-lens#168: Implements eager-loading behavior described in this release’s skill guidance and release content.

Suggested reviewers: syedmustafan

Merge Risk: Low · up to 394c5

Users may run a different analyzer version than the documented 1.13.6 release, producing inconsistent scan behavior.

Finishing Touches

📝 Generate docstrings

• Commit to this branch
• Create a new PR

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}