[JustinKovacich]

Writing pass over crates/uds_protocol to the "Writing for the reader" standard (CONTRIBUTING.md, #84, which this PR is stacked on). Prose only: rustdoc, comments, the Cargo.toml feature comments, the fuzz targets and one word in the README (whose intra-doc links are unchanged; it is also the crate doc). No code line changes: the non-comment lines in the diff are nine renamed test functions, three trailing comments removed from code lines, and two reworded assertion messages.

Before / after (prose lint): hard-fail phrases 27 / 0 (one "load-bearing", ten first-person, sixteen "deliberately"); history 54 / 0; incident 5 / 0; editor instructions 2 / 0; # Why headings 1 / 0; em-dash 108 / 0; comment blocks of 12+ lines 45 / 9. The nine that remain are Error::negative_response_code's example and three-code mapping, the six RequestFileTransfer mode tables with their link definitions, DtcStatusMask's example and bit table, and AddressAndLengthFormatIdentifier's example.

Test renames (nine, each a sentence of 12 or more words; none in a CI filter; every prop_ name kept): for example a_tester_can_still_request_an_unmodeled_read_dtc_information_sub_function is now try_reserved_builds_an_unmodeled_sub_function, and the SERVICES doc that named the_variants_that_are_not_a_2020_request_service_report_no_sub_function follows its rename.

Fixed while reading: fuzz_roundtrip.rs said Request has no PartialEq (it derives it); a doc comment describing whole_records was attached to encode_by_dtc_number (moved); the upload/download $repr doc called AddressAndLengthFormatIdentifier crate-private (it is public). The two "this doc is the published schema description" notes on the repr structs are now // comments, so they are not published.

Removed, and the reasoning it held (this PR is its record; nothing is filed elsewhere):

• Why Request::is_positive_response_suppressed returns Option<bool>: a wrong false for a vendor request made a session layer start tP_Client, time out and retry twice. One sentence of this stays. Candidate Rationale: for the need that owns SPRMIB reporting.
• Why has_sub_function answers None for the withdrawn 0x83; why allowed_nack_codes is a floor (clause 9.4; 0x78 is in no table; a whitelist would reject every ResponsePending), kept to one sentence; the full account of which codes negative_response_code never returns and why.
• modeled_frames sets the SPRMIB on every sub-function service, and the dispatch test once asserted only !is_empty(); assert_encode_size_agrees is not byte coverage (a swapped field pair once passed the suite).
• The crate-wide lint derogations' site counts (162 production sites, 281 test sites).
• The incidents behind some forty test comments across the service modules: the communicationType byte matched whole; manufacturer DTCSettingType ranges rejected; a spurious powerDownTime byte; 0x82/0x8A falling through to IsoSaeReserved; the missing DTCFormatIdentifier; the one-byte-too-large shortfall; the non-terminating partial-tail iterator; three missing RequestFileTransfer codes; NamePayload's mode field winning over the variant; TesterPresent normalising reserved bytes to 0x00; a 3-byte WDBI decoding to an empty record; 0x7F normalisation of unmodeled SIDs. Each test now states the behaviour it pins and its ISO citation.
• The three OpenAPI defects that shipped before tests/openapi_schema.rs existed; the five serde escape routes tests/public_api.rs closes, described in the present tense now; the mirror-struct cost (about a hundred lines for a one-field invariant); the utoipa-without-serde build that was tried.
• tests/spec_conformance.rs: the earlier comment's wrong claim about Tables 31, 38 and 47, and which tables' byte columns the markdown conversion garbled.
• Smaller: why DataFormatIdentifier::new uses explicit range checks (? in a const fn); TryFrom<u32> over masking; new_with_alfid for fixed-ALFID bootloaders (one sentence stays); a review probe that sent 27 01; automotive-wire-codec finding SE-1.

Issue URL

No issue; the pass follows the standard agreed in #84.

Testing

• python3 tools/prose_lint.py crates/uds_protocol --fail --show 50 exits 0 with every hard-fail row at 0; CI's Comment prose job now runs --strict crates/uds_protocol.
• cargo fmt --all --check; cargo clippy -p uds_protocol --all-targets --all-features -- -D warnings; cargo nextest run -p uds_protocol --all-features (298 passed); cargo doc -p uds_protocol --all-features --no-deps with -D rustdoc::broken_intra_doc_links -D warnings.
• git diff -U0 ... | grep -vE '^[-+]\s*(//|#|\*|/\*|$)' prints only the renamed test functions, the three code lines whose trailing comments went, and the two assertion strings.

🤖 Generated with Claude Code