docs(contributing): correct four small claims

[michen00]

What changes, and why

Four small claims in CONTRIBUTING.md, each recomputed.

:33 — what tests/test_unwrap.py covers. "covers only what a corpus cannot describe, because no other implementation shares it: argument handling, file discovery, encoding failures, exit codes." Its own module docstring gives five items, adding the transcript skip. And the stated reason is wrong for part of the file: the matcher-level tests are shared — src/cli.rs and src/ignore.rs carry the same vectors as Rust unit tests, because a matcher is where a hand-written port diverges in a way no document would catch. The sentence now separates the two groups and says why the second exists.

:43 — the version floor. Quoted as requires-python = ">=3.10"; the manifest is requires-python = '>=3.10' (pyproject.toml:21). A literal search for the quoted line found nothing.

:63 — the notebook's figures. "Every figure in it is computed by the cell above it, so no number is written into the prose." No measurement is, which is the property worth claiming and the one the cells enforce. Counts of what a cell does are spelled out in the markdown — "times three programs that do almost nothing", "six of the eight measurements would fall within the first quarter", "times two runs" — and those track the cell's own constants. The absolute version promised something the notebook does not deliver, which matters because this sentence is what a reader relies on when deciding whether a prose number can be trusted.

:73 — the nbconvert timeout. "The default timeout is 30 seconds per cell, and several cells need longer." There is no per-cell default:

$ uv run python -c "from nbclient.client import NotebookClient; ..."
  nbconvert 7.17.1 / nbclient 0.11.0
  NotebookClient.timeout default: None

nbconvert hands execution to nbclient, and the 30-second default belonged to nbconvert 5.x before that. So --ExecutePreprocessor.timeout=1800 is a ceiling where there was none, not a short limit raised — still worth passing, for a different reason than the one given, and a contributor was otherwise told to expect a hung cell to abort in 30 seconds when it would run forever.

Corpus

The corpus is the specification, and both implementations answer to it. Tick what applies.

• This changes no behavior the corpus specifies.
• This changes what gets joined, and a case in corpus/ pins the new behavior. The case was written first and failed first.
• The change makes the tool join more than it did. The section above says what it will not eat.

Checks

• make check passes, or make test does and this touches no Rust.

make tidy, plus unwrap-markdown-prose-py --fail-on-change CONTRIBUTING.md. One Markdown file.

[qodo-code-review]

Code Review by Qodo

🐞 Bugs (0) 📘 Rule violations (0) 📎 Requirement gaps (0)

Great, no issues found!

Qodo reviewed your code and found no material issues that require review

Tip of the day

💡 Did you know, you can add REVIEW.md to your repo root and Qodo follows it on every PR

More tips ↗ | Customize Qodo ↗ | Qodo docs ↗

[qodo-code-review]

PR Summary by Qodo

Correct contributing guide claims about tests and benchmarks

📝 Documentation 🕐 Less than 5 minutes

AI Description

• Clarifies Python-only and shared matcher-level test coverage.
• Corrects the Python version declaration and notebook measurement guarantees.
• Explains nbclient’s unlimited default timeout and the explicit execution ceiling.

High-Level Assessment

Directly correcting the inaccurate statements in the contributing guide is the appropriate approach. Code, tests, and configuration already express the underlying behavior, so changing them instead would be unnecessary and misleading.

Files changed (1) +4 / -4

Documentation (1) +4 / -4

CONTRIBUTING.mdCorrect test, version, benchmark, and timeout guidance +4/-4

Correct test, version, benchmark, and timeout guidance

• Distinguishes Python-specific test coverage from matcher vectors shared with Rust and adds the omitted transcript-skip case. It also matches the manifest’s exact Python requirement syntax, narrows the notebook guarantee to measured figures, and explains that the explicit nbclient timeout adds a ceiling where none exists by default.

CONTRIBUTING.md

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 87.40%. Comparing base (9e6280e) to head (183d6ee).

Additional details and impacted files

@@           Coverage Diff           @@
##             main      #43   +/-   ##
=======================================
  Coverage   87.40%   87.40%           
=======================================
  Files           3        3           
  Lines         691      691           
=======================================
  Hits          604      604           
  Misses         87       87           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

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

[Copilot]

🟢 Approval recommended

Changes are documentation-only and the updates are consistent with the referenced repository sources, with only a minor wording nit noted.

Pull request overview

This PR updates CONTRIBUTING.md to correct four previously inaccurate or over-absolute documentation claims, aligning contributor guidance with what the repository actually does (tests, version floor declaration, benchmark notebook guarantees, and nbconvert timeout behavior).

Changes:

• Clarifies what tests/test_unwrap.py covers vs. what the corpus specifies, including the transcript-skip and matcher-level parity pins.
• Corrects the quoted requires-python declaration to match pyproject.toml.
• Refines benchmark notebook statements to distinguish measured figures from prose counts, and updates the nbconvert timeout explanation; adds guidance to re-execute the notebook when lint fixes touch code cells.

File summaries

File	Description
CONTRIBUTING.md	Corrects and tightens several contributor-facing claims about tests, version floor, and benchmark notebook execution guarantees.

Review details

• Files reviewed: 1/1 changed files
• Comments generated: 1
• Review effort level: Lite

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.