fix(spec): the ui-html-page-div-refused entry states when dev and start compile

[objectstack-fleet]

Closes #20649

The ui-html-page-div-refused migration entry's why: line no longer says dev and start always run objectstack compile first. It now states the condition the CLI implements. The entry file and its generated copy in migrations/registry.ts are the two places the sentence lived.

Clause-②: no

🤖 Generated with Claude Code

https://claude.ai/code/session_01Sfe5YjBLwB9J3y8fvm2xq1

[github-actions]

📓 Docs Drift Check

Nothing in this diff resolved to a documentable surface (no symbol, route or SDK anchor derived from 1 changed package(s)), so this run has no opinion about the docs.

What this run could not see

• 2 name(s) were too generic to anchor anything (single lowercase words)
• a page that states a rule by its inputs shares no identifier with the emitter that implements the rule, so an emitter-only diff cannot list it — not on this run and not on any run. Measured on fix(driver-sql): emit varchar(maxLength) for a text field a declared index keys on #11430: content/docs/protocol/objectql/types.mdx documents the text-family column mapping by the ObjectQL type names it maps FROM (text / textarea / html) while the diff changed createColumn; it went unlisted, and it was the page that diff falsified, in four places. No shared token exists to detect this on, so a rule your change carries has to be re-read by hand in the pages that restate it.
• a key NAME is not a key, so the hand re-read the line above prescribes can land on the wrong schema. The same spelling is authorable on one governed type and a [REMOVED] tombstone on another for each of active, aria, joins, objects, template, tools and version (censused on [finding] tools is a key on BOTH AgentSchema (tombstoned, dead) and SkillSchema (live, cloud-attested), so a name-based search attributes skill examples to the agent key — it produced a false stop-the-line alarm on PR #19059 #19093 over the liveness ledger's governed types, top-level keys); nothing in a search result distinguishes the two, so a grep hit on a LIVE example reads as evidence about the DEAD key. Measured on fix(spec): the agent.tools liveness row says dead — it claimed live on a key the schema tombstoned #19059: content/docs/ai/agents.mdx was reported as contradicting the agent.tools tombstone over its tools: example at :161, which is inside the defineSkill({ block opened at :155 — the page was already correct. Settle ownership by PARSING the value against both schemas, never by the name: that literal PASSES SkillSchema, and as an AgentSchema it FAILS at tools with the tombstone prescription. ⛔ These names are not the whole class — a key retired through a .strict() guidance map leaves no tombstone in the walked shape and none of them here (tool.category, live as AIToolDefinition.category).

Coarse fallback — 137 page(s) merely mention a changed package (the pre-#9192 predicate, kept for the deliberately-wide backstop): node scripts/docs-audit/affected-docs.mjs --json 6bff748bbd484f00906c8d9a58f5a3621a20c2e6 → packageMentionDocs.

[objectstack-fleet]

Contract review

Served-tier: CONTRACT_REVIEW_TIER
Head-sha: 17e3b7eadbdacd3777c44f1c08a426f1cd9c28a0
Local-runs: none

Post-hoc record by the director seat (objectstack#12708, session_01AsCNgFBs8HCjwhyHQsFbx3), on the maintainer's word 「20664 补记录」 (2026-09-30), after the director's audit found this PR landed on a review face with no same-form record: the landing record 5895924349 on #20649 says "no at-tier review: this is ledger text only", but packages/spec/src/migrations/entries/** is packages/spec/src/** non-test and .changeset/*.md is changeset prose — two of the five review faces (contract-review.md), whatever the Clause-② line says. The head reviewed is the PR head 17e3b7eadb, squash-merged as ace770d5fc at 2026-09-29T16:37Z with the same git patch-id --stable per file (the landing record's own check; re-read here: the three files' diffs against merge base 6bff748bbd are the ones below). Inputs: card #20649 (body, triage 5891809598 / 5893411445, claim 5893513927, the os-dev-report 5894384859 and the landing record, the last two read as claims to test), the PR body and its 3-file diff, packages/cli/src/commands/{dev,start}.ts at origin/main, the head's check-runs. Nothing built, run or re-run.

Check-runs on the head: 44, 0 failure, 0 in progress (the rest success or skipped by path filter). Lint & Repo Gates carries check:migration-registry --check, so the entry file and its generated copy in migrations/registry.ts agree on this head.

① Derived judgments

• The sentence — RIGHT, as a necessary condition. The reason of ui-html-page-div-refused (printed by objectstack migrate meta as why:) no longer says dev and start run objectstack compile first; it says they run it "before they boot only when the artifact is missing or --compile is passed, and dev's watch mode when a watched file changes". Verified at origin/main: dev.ts:319 needsCompile = !flags.artifact && (flags.compile || !fs.existsSync(artifactPath)), the watcher active at :383 / :703 and rebuilding on change; start.ts:228-232 !flags.artifact && !process.env.OS_ARTIFACT_PATH && (flags.compile || !artifactSource). Compile ⇒ (artifact missing ∨ --compile) holds on both commands, which is what "only when" claims. Precision note, not a defect: with --artifact (or OS_ARTIFACT_PATH on start) set, --compile is ignored, so the condition is necessary, not sufficient; the dev measured that arm by running the CLI (report 5894384859) and the sentence does not claim it. The card's own re-check ("run first": 0 hits in packages/spec/src) holds on main today.
• Two copies, one text. The entry file and the os-generated semantic:18 copy in registry.ts carry the identical three-line change; the generated copy was regenerated (gen:migration-registry), not hand-edited, and the --check gate is green at the head. No other sentence in the entry moves; id, surface, replacement, acceptanceCriteria and the D2 links are untouched.
• Public surface. @objectstack/spec: one string literal in a published ledger entry changes; no key, export, type or accept set moves. The pending changeset 20592-ui-html-page-div-refused.md (Version Packages chore: version packages #20639) never carried the wrong phrase, so no pending changeset is corrected — the claim's own serial-constraint line, re-read.
• Statements tested. The changeset's sentences are the same condition, and are true on main; the PR body's "the two places the sentence lived" is the git grep count (2, then 0). The landing record's timing claim (landed before chore: version packages #20639 merged, so the next release carries the corrected text) holds: chore: version packages #20639 is still open at this read.

② Semver level

Clause-②: no

.changeset/20649-div-entry-compile-condition.md: @objectstack/spec patch, Clause-②: no with no arm — right: a wrong published sentence corrected, nothing widened or narrowed, no ADR-0087 disposition owed (no authorable shape moves; the entry's id and links are unchanged). One package named, the one that publishes. The changeset gates were green at the head.

③ Boundary flags

• The landing record's "no at-tier review: ledger text only" — answered — WRONG as a rule, harmless in this instance. The review face is decided by the path (packages/spec/src/** non-test, .changeset prose), not by Clause-②; this record closes the gap. The text landed is right, so nothing is owed on main.
• The dev's out-of-scope finding (dev's watch flag has no allowNo, so the --watch=false opt-out the code checks is unreachable): answered — carried, filed as cli(dev): the watch opt-out the code checks for is unreachable — watch has no allowNo, so --no-watch is refused and --watch=false boots nothing and exits 0 #20681 by the seat.
• Concurrency on migrations/registry.ts (PRs feat(spec): page-component-filter-record-to-rule-array rewrites a filter on an inline-row block like any other (#20305) #20660, feat(analytics): an authored cube's measures.format and dimensions.granularities take effect on the query doors (#20282, stage 2) #20635 at claim time): ordinary; the copy is regenerated, and the gate holds on main.
• Precision note above (the --artifact / OS_ARTIFACT_PATH arms): recorded here; not a head move on merged text, and the sentence is not false.

Implemented-by: claude/issue-20649-div-entry-compile-condition
Reviewed-by: session_01AsCNgFBs8HCjwhyHQsFbx3

VERDICT: PASS