[JarryShaw]

Please follow the guide below

• Searched for similar pull requests
• Followed the coding style (make pylint, make mypy, make isort) — ran pylint/mypy/isort with the Makefile flags on the changed files only; no new findings
• make test passes, and a test case covers the change — the modules listed below only, not the full suite
• Added a changelog entry — N/A, added centrally

What is the purpose of your pull request?

Tick the commit type your subject line carries.

• fix — corrects a defect
• feat — adds a feature
• perf — changes performance, not behaviour
• refactor — changes neither behaviour nor performance
• test — tests only
• docs — documentation only
• ci — workflows or build tooling
• release — bumps the version or rolls up a distribution
• chore — anything else

---

Description of your pull request and other information

Closes #1185. _make_opt_mpl (HOPOPT and IPv6_Opts) now takes S from the parsed seed_type (or a new seed_type keyword), and infers it only when neither is given. For S=0 the seed is the IPv6 source address and is not written. Before, inferring from bit_length() raised on that IPv6Address, and a 64/128-bit seed with a small value was rebuilt as S=1. A missing or oversized seed raises ProtocolError. The MPLOption.seed_id annotation now includes IPv6Address.

Probe (from_data on every HOPOPT/IPv6_Opts layer in examples/captures/*): before, 2 of 28 raise AttributeError (options-ipv6.pcap frames 9 (HOPOPT) and 23 (IPv6_Opts) by info.number, in a capture regenerated from this PR's base 20438e3ec). After, 0 of 28 raise, and both frames rebuild byte-exactly.

Tests: new tests/protocols/internet/test_mpl_seed_rebuild_unit.py covers S=0–3 byte-exact rebuild, S=0 inside IPv6 and the rejection cases. Without the fix: 12 failed, 4 passed (16 subtests in total). With the fix: 4 passed, 18 subtests passed. test_option_roundtrip_unit 6 passed (no MPL EXPECTED_FAILURES entries), test_ipv6_ext_unit 30 passed, test_ipv6_extension_runtime 3 passed, test_option_coverage_runtime 3 passed, test_declared_length_roundtrip_runtime 6 passed, tests/project 389 passed / 1 skipped. util/run_unittest_leg.py protocols/internet: 476 tests, 0 failures, 0 errors.

[github-actions]

Coverage: 88.89% (unit tier, Python 3.14, e828f35ac, Unit Tests run success)

Package	Statements	Missed	Branches	Partial	Cover
pcapkit (top level)	104	4	20	4	93.55%
pcapkit/const	18757	1034	2342	845	90.24%
pcapkit/corekit	1878	91	580	22	94.34%
pcapkit/dumpkit	142	0	42	0	100.00%
pcapkit/foundation	2442	102	852	39	94.38%
pcapkit/interface	112	7	40	5	92.11%
pcapkit/protocols	15731	190	3998	165	98.17%
pcapkit/toolkit	487	65	144	1	85.42%
pcapkit/utilities	429	4	122	4	98.55%
pcapkit/vendor	4409	2359	1006	158	42.84%

Per-file detail: the coverage-html artifact of this run.

[JarryShaw]

Cross-review verdict on 4fc32b635: NEEDS CHANGES (ran on Sonnet; author Opus). One wrong citation; the code is good. Labelled review: needs-changes.

• Change needed: the new test's module docstring cites :rfc:7731#section-2.1``. The S field is defined in RFC 7731 §6.1 (MPL Option). I confirmed this against the RFC text.
• Confirmed:
  • Seed types per S: None or IPv6Address for S=0, int for S=1–3.
  • Every S rebuilds byte-exactly, standalone and inside IPv6, for both classes, including small S=2/3 seeds.
  • seed_type follows the module's *_default/_namespace/_reversed pattern, and out-of-range values raise ProtocolError.
  • The HOPOPT and IPv6_Opts copies behave identically.
  • Without the fix, 12 tests fail.
  • The protocols/internet leg runs 476 tests, 0F/0E.
• Captures: both MPL frames in options-ipv6.pcap now rebuild as 11006d0200000000. Their numbers depend on the indexing: 8 and 21 by info.number, against the body's New distribution [0.13.1] #9/Fixed minor bugs in Makefile #23. The 28-layer total across all captures was not re-run.
• Out of scope: _make_opt_smf_dpd rebuilds H-DPD as I-DPD. Filed as fix(ipv6-opts): _make_opt_smf_dpd overwrites the parsed DPD mode, so H-DPD rebuilds as I-DPD #1190, blocked on this PR.
• Nit: seed_type=1 with an IPv6Address seed raises with a misleading "too large" message.

[JarryShaw]

Cross-review verdict on e828f35ac: GOOD TO GO (ran on Sonnet; author Opus). Labelled review: good-to-go.

• Interdiff from 4fc32b635: exactly the requested one-line fix. The test docstring now cites 7731#section-6.1. I re-ran git diff 4fc32b635 e828f35ac myself.
• Tests: the new test module passes on the new head.
• Frame numbers: the PR body's info.number 9 (HOPOPT) and 23 (IPv6_Opts) are correct. The reviewer's earlier 8/21 came from a stale copied capture; the regenerated one has 28 HOPOPT/IPv6_Opts layers, none raising, and both MPL layers rebuild as 11006d0200000000.
• Unchanged from round 1: every other check carries over, since the source files are identical.