docs(register): C-86 and C-87 — the residuals ADR-016 and ADR-017 leave behind

[Polichinel]

Answering a direct question: "are all risks registered?" — they were not. Two.

Both were described in the ADRs and tracked nowhere. A decision record states a position; the register is what carries a live risk.

C-86 — the release path depends on two other repositories, with no way past a red build

ADR-016 accepted that coupling partly on a justification measurement killed: protect_main lists zero bypass actors, so administrator status confers no exemption, and there is no classic branch protection either. That is withdrawn in the ADR; this entry is the risk it was papering over.

The interesting part is that it is latent. No status check is required on main today, so a red build blocks nothing and the coupling costs nothing.

The moment C-81's enforcement half is fixed — correctly, and it should be — this becomes live and unbypassable in the same change. Two open items that each look sensible on their own combine into something neither of them says. C-81 now says it.

The rate is not hypothetical: views-appwrite reports five registry editions in four days, four observation-only, each of which would have blocked a release. views-appwrite#76 filed to make the obligation-carrying distinction machine-readable — the preferred resolution, and one that seat volunteered.

C-87 — the label will be checked against a declaration; nothing will check the declaration against the consumer

ADR-017 §8 states it plainly. It belongs in the register because the second half is work in repositories this project does not control — views-faoapi#379 has a willing owner, views-crafdapi#39 is blocked on that partner's data contract — and because for a third-party-operated private API it may never exist at all.

Also records the ordering constraint that a reviewer's catch produced: the source-reading check must survive until the consumer-side check lands, or there is a green window proving only that two values this platform authored agree with each other.

Also filed while checking

views-appwrite#76 — the pin-by-contract-version option. It was recorded in ADR-016 §7a as "not adopted here, needs the upstream side first" and tracked nowhere, which is how an idea dies in a paragraph. views-appwrite#45 is adjacent but different: that one is about readers ignoring [meta] version; this is about which changes should oblige a consumer to act.

Register 85/12/73 → 87/14/73.

ruff check .        # clean
pytest -q           # 399 passed, 40 xfailed

🤖 Generated with Claude Code