These goldens were recorded on a machine with personal skills under ~/.dsh and
~/.agents, so every one of them pinned a context-injection row a clean runner
cannot produce — which is why the browser lane failed on CI and passed
locally. master's scaffold fix confines skill-local's host-level roots to the
temp world; re-recording against it removes the row and nothing else.
seeded-history keeps its row: that scenario seeds the event itself.
Two lifecycle holes the registry and the consumer helper left open.
`AdapterRegistrationHandle.replace` had no liveness guard: after the handle's
disposer ran, a replace put routes back into the registry with nothing left to
release them, so the adapter leaked permanently. `owned` being empty cannot
carry that fact, because `replace([])` is the legal empty-section state, so the
disposer records it explicitly.
`installSettingsSection`'s watcher lacked the guard its own disposer carries:
a stored change landing while the consumer unloads reached `onChange`, which
re-registers routes against a fiber whose resources are being released.
Also documents `withFileLock` in the atomic-write README (it claimed one
export), records the age-based lock takeover as a known limitation, and lists
ctx.settings and ctx.credentials in the architecture capability table.
The base's ReadResultView now requires `offset` (persisted so an empty window
still knows its start). The connection fixture's read sample and read-card.spec's
resultRead helper supply it (the window's first line, 41).
Generator correctness stays pinned by its fixture suites and the
byte-for-byte catalog reproduction test; per-file coverage put
whole-workspace compiler analysis under v8 instrumentation — the
coverage lane's longest tail. Widen the existing three-file exclude to
the package's src.
The nested read of a missing file surfaces its error state as 失败 in the
code-dispatch row summary, a consequence of this PR's read-card error
presentation. The golden predated it (never re-recorded while the PR was
conflicting); refreshed keyless via DSH_SNAPSHOT=refresh over the existing
session.jsonl.
Keep only groups with another path in the active round. This bounds skewed sampling by paths visited instead of rescanning every singleton for every late-group item.
The Note's pre-release-stance link used the wrong depth and target
(../../../CLAUDE.md); point it at ../../../../AGENTS.md with the section
anchor so verify-md-links passes. The presentResult decline test's meta
lacked the now-required offset, so it declined at meta narrowing instead of
exercising the content-shape decline (read.ts:181-183); add offset back.
