Step 6 of "Adding a new device" still said to add `- <device>` to the matrix in
.github/workflows/master.yml, and called that matrix "the *only* build registry".
Neither has been true since #120 moved the matrix into .github/scripts/ci-matrix.py,
where it is derived from the tree rather than listed. master.yml has no device list
at all.
Not a hypothetical staleness: #111 (Tapo TC70 v3) and #108 (KC110) each append a line
to a block that no longer exists, so both are merge conflicts rather than the one-line
changes they were written as. Their authors did the documented thing.
So step 6 is now the opposite instruction -- there is nothing to add, and adding it is
the mistake. It also names NOT_BUILT, since that is the one list a contributor could
reasonably go looking for, and says to leave it alone.
Two adjacent claims went stale with the same change: the master.yml bullet described a
cron-and-dispatch workflow with no pull_request trigger, and nothing mentioned
ci-matrix.py at all -- including that --self-test runs in the Select devices job and
will fail a PR that puts the tree and the selector out of step.
A second correction, caught by Qodo in review: "opening the PR is what gets the device
built" holds only for a non-draft PR. classify() returns needs_build=False for a draft
and self_test() asserts it; master.yml lists ready_for_review in its triggers precisely
because a PR opened as a draft would otherwise never build once marked ready. Left as
written, a contributor could open a draft, see no build, and read that as the device
being fine. The ci:full label is documented for the same reason -- it is the other
contributor-facing lever over the matrix.
Document the builder overlay model (per-device deltas layered onto a
fresh OpenIPC/firmware clone), the build flow in builder.sh, device
directory anatomy, the package/ and devices/common/ conventions, the CI
workflows (including the moving-ref cache caveat), and a step-by-step
"Adding a new device" walkthrough.
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
